Functional Anthology of Intrinsic Disorder. 1. Biological Processes and

Functional Anthology of Intrinsic Disorder. 1. Biological Processes and Functions of Proteins with Long Disordered Regions Hongbo Xie,† Slobodan Vucetic,† Lilia M. Iakoucheva,‡ Christopher J. Oldfield,# A. Keith Dunker,# Vladimir N. Uversky,*,#,§ and Zoran Obradovic† Center for Information Science and Technology, Temple University, Philadelphia, Pennsylvania 19122, Laboratory of Statistical Genetics, The Rockefeller University, New York, New York 10021, Center for Computational Biology and Bioinformatics, Department of Biochemistry and Molecular Biology, Indiana University, School of Medicine, Indianapolis, Indiana 46202, and Institute for Biological Instrumentation, Russian Academy of Sciences, 142290 Pushchino, Moscow Region, Russia Received August 2, 2006

Identifying relationships between function, amino acid sequence, and protein structure represents a major challenge. In this study, we propose a bioinformatics approach that identifies functional keywords in the Swiss-Prot database that correlate with intrinsic disorder. A statistical evaluation is employed to rank the significance of these correlations. Protein sequence data redundancy and the relationship between protein length and protein structure were taken into consideration to ensure the quality of the statistical inferences. Over 200 000 proteins from the Swiss-Prot database were analyzed using this approach. The predictions of intrinsic disorder were carried out using PONDR VL3E predictor of long disordered regions that achieves an accuracy of above 86%. Overall, out of the 710 Swiss-Prot functional keywords that were each associated with at least 20 proteins, 238 were found to be strongly positively correlated with predicted long intrinsically disordered regions, whereas 302 were strongly negatively correlated with such regions. The remaining 170 keywords were ambiguous without strong positive or negative correlation with the disorder predictions. These functions cover a large variety of biological activities and imply that disordered regions are characterized by a wide functional repertoire. Our results agree well with literature findings, as we were able to find at least one illustrative example of functional disorder or order shown experimentally for the vast majority of keywords showing the strongest positive or negative correlation with intrinsic disorder. This work opens a series of three papers, which enriches the current view of protein structure-function relationships, especially with regards to functionalities of intrinsically disordered proteins, and provides researchers with a novel tool that could be used to improve the understanding of the relationships between protein structure and function. The first paper of the series describes our statistical approach, outlines the major findings, and provides illustrative examples of biological processes and functions positively and negatively correlated with intrinsic disorder. Keywords: intrinsic disorder • protein structure • protein function • intrinsically disordered proteins • bioinformatics • disorder prediction

Introduction Among other objectives, computational biology aims to enable an understanding of the relationships between the primary sequence, the higher order structure, and the function of proteins. Each protein function is generally thought to originate from a specific three-dimensional (3D) structure. * Correspondence should be addressed to: Vladimir N. Uversky, Department of Biochemistry and Molecular Biology, Indiana University School of Medicine, 635 Barnhill Drive, MS#4021, Indianapolis, IN 46202. Phone, 317278-9194; fax, 317-274-4686; e-mail: [email protected]. † Temple University. ‡ The Rockefeller University. # Indiana University. § Russian Academy of Sciences.

1882

Journal of Proteome Research 2007, 6, 1882-1898

Published on Web 03/29/2007

Formulation of this view began more than 100 years ago with the lock-and-key model proposed by Fischer.1 More than 70 years ago, Wu2 and, slightly later, Mirsky and Pauling3 equated denaturation with loss of specific structure. The dependence of function on 3D structure was accepted by the time of the protein folding studies of Anfinsen and colleagues.4 The flood of protein 3D structures determined by X-ray diffraction and by nuclear magnetic resonance (NMR) spectroscopy has overwhelmed alternative concepts.5 In contrast to the dominant view given above, proteins for which intrinsic disorder is required for function have been reported in the literature for many years. By “intrinsic disorder” we mean that the protein (or protein region) exists as a structural ensemble, either at the secondary or at the tertiary 10.1021/pr060392u CCC: $37.00

 2007 American Chemical Society

Functions of Intrinsic Disorder. Part 1

level. Thus, both extended regions with perhaps some elements of secondary structure and collapsed (molten globule-like) domains with poorly packed side chains are included in our view of intrinsic disorder.6 More detailed analysis of extended disordered proteins/regions revealed that they can be further divided into two groups, random coil-like and pre-molten globule-like conformations.7 Recently, more than 150 proteins have been identified as containing functional disordered regions, or being completely disordered, yet performing vital cellular roles.8,9 Twenty-eight separate functions were assigned to these disordered regions, including molecular recognition via binding to other proteins or to nucleic acids.8,10 A complementary view is that functional disorder fits into at least five broad classes based on the mode of disordered protein/region action.10 Obviously, for these proteins, the predominant structure-function paradigm is insufficient, suggesting that a more comprehensive view is needed.11 In fact, a new paradigm was recently offered to elaborate the sequence-to-structure-tofunction scheme in a way that includes the novel functions of disordered proteins.6,7,12 The complex data supporting this revised view were summarized in “The Protein Trinity” hypothesis, which suggested that native proteins can exist in one of three states, the solid-like ordered state, the liquid-like collapsed-disordered state, or the gas-like extended-disordered state.12 Function is then viewed to arise from any one of the three states or from transitions between them. Later this paradigm was extended to “The Protein Quartet” model to include one more extended-disordered conformation, the premolten globule state.7 For structured proteins, that is, proteins that form crystals without partners or have ordered globular forms without partners in NMR experiments, we will use the terms “structured”, “intrinsically ordered”, or just “ordered”. Recent studies revealed that many proteins lack rigid 3D structure under physiological conditions in vitro, existing instead as highly dynamic ensembles of interconverting structures. Indeed, the literature on these proteins, known as intrinsically disordered, natively unfolded, or intrinsically unstructured, has virtually exploded during the past decade.7,13 This literature explosion is consistent with bioinformatics studies predicting that about 25-30% of eukaryotic proteins are mostly disordered,14 that more than half of eukaryotic proteins have long regions of disorder,14,15 and that more than 70% of signaling proteins and the vast majority of cancerassociated proteins have long disordered regions.16 As it has been already mentioned, despite the fact that intrinsically disordered proteins fail to form fixed 3D structures by themselves under physiological conditions, they carry out numerous important biological functions.6-11,13,16-22 Intrinsically disordered regions are typically involved in regulation, signaling, and control pathways in which interactions with multiple partners and high-specificity/low-affinity interactions are often involved.21-23 Furthermore, sites of post-translational modifications (acetylation, hydroxylation, ubiquitination, methylation, phosphorylation, etc.) and proteolytic attack are frequently associated with regions of intrinsic disorder.6,16 Given the high frequency of intrinsically disordered proteins and their crucially important functions, a curated Database of Disordered Protein (DisProt) has been recently initiated.24 This database provides structure and function information about proteins that lack a fixed 3D structure under putatively native conditions, either in their entireties or in part.24 Nevertheless, the phenomenon

research articles of intrinsic disorder in proteins is still severely underappreciated; not a single biochemistry textbook discusses these proteins.25 There is a large gap between the number of proteins with experimentally confirmed disordered regions and the actual number of such proteins in nature. Although studies of functional properties of known disordered proteins are helpful in revealing the functional diversity of protein disorder, they are bound to provide only a limited view. In this study, we propose a statistical approach for comprehensive study of functional roles of protein disorder. This approach relies on use of the VL3E26 predictor that is currently the most accurate predictor of long disordered regions with estimated accuracy of above 86%.26 The high accuracy of VL3E ensures that most disordered regions could be successfully detected, and only a small fraction of ordered regions are incorrectly labeled as disordered. The VL3E predictor was applied to over 200 000 Swiss-Prot27 proteins, many of which were annotated with one or more functional keywords. Then, the disorder- and ordercorrelated functions were detected as those that are overrepresented by proteins predicted to have long disordered regions (>40 amino acid residues) in comparison with a random selection of proteins with the same length distribution. The proposed approach ensures that adverse effects of sequence redundancy and sequence length are eliminated. Disorder predictors were previously used to analyze functions of disordered proteins. For example, it was shown that a large fraction of cancer-related proteins are likely to be disordered.16 Another study28 demonstrated that many processes in yeast are related to protein disorder. The current study provides a comprehensive analysis of disorder-related functions by using a much larger set of proteins (i.e., the entire Swiss-Prot database). Given a list of functions positively and negatively correlated with disorder, we performed an extensive literature survey to find experimental evidence supporting the findings. We were able to find at least one illustrative experimentally validated example of functional disorder/order for a large majority of functional keywords. This work opens a series of three papers dedicated to the discovery and description of protein functions and activities that are positively and negatively correlated with long disordered regions. The first in the series, this paper deals with the description of the statistical approach used here and delineates the major results of the application of this tool for the analysis of over 200 000 proteins from the Swiss-Prot database. This paper also provides illustrative literature examples related to the Swiss-Prot keywords associated with the biological processes and functions positively and negatively correlated with intrinsic disorder. The second paper of the series portrays keywords related to the cellular components, domains, technical terms, developmental processes, and coding sequence diversity associated with long disordered regions,29 whereas keywords correlated with ligands, post-translational modifications, and diseases associated with long disordered regions are the topic for the last paper of the series.30 The overall result is that this series of papers represents a functional anthology of intrinsic disorder that includes both the results of our bioinformatics analysis and illustrative literature examples for the majority of functional keywords possessing strongest positive or negative correlation with the intrinsic disorder prediction. Journal of Proteome Research • Vol. 6, No. 5, 2007 1883

research articles

Xie et al.

Materials and Methods Data Set. The data set for analysis was constructed using the Swiss-Prot database (release 48, 2005) containing 201 560 proteins.27 In this study, we used the 196 326 proteins with length longer than 40 amino acid residues. Each protein in Swiss-Prot is annotated with keywords that describe its functional or structural properties. Out of the 874 keywords used by Swiss-Prot, 710 were associated with at least 20 proteins. Swiss-Prot is statistically redundant, as it contains a large number of homologous proteins with highly similar sequences.31 Ignoring the redundancy would significantly bias statistical inference. To reduce redundancy, TribeMCL32 was applied to cluster the protein sequences from Swiss-Prot into families. TribeMCL uses the Markov clustering algorithm for the assignment of proteins into families based on the similarity matrix generated from the all-against-all BLASTp33 comparison of sequences. It is able to produce high quality families despite presence of multidomain proteins, peptide fragments, and promiscuous domains.32 The obtained BLAST profiles were imported into TribeMCL software package (http://micans.org/ mcl/), and clustering was performed with all parameters set at default. As a result of the application of this redundancy reduction procedure, the sequences were grouped into 27 217 families. Predicting Long Disordered Regions in Proteins. Previous studies suggested that in comparison with ordered sequences, disordered sequences tend to have lower aromatic content, higher net charge,17,34-36 higher values of the flexibility indices, greater hydropathy values,34,36 and lower sequence complexity.37 Following these observations, the VL3E predictor26 was developed using 162 long (>30 residues) disordered regions from a nonredundant set of 152 DisProt proteins24,38 and 290 completely ordered proteins. The predictor consists of an ensemble of neural network classifiers, and it achieves ∼87% cross-validation accuracy on balanced data with equal number of ordered and disordered residues. We used the VL3E predictor to predict Swiss-Prot proteins with long disordered regions. Each of the 196 326 Swiss-Prot proteins was labeled as putatively disordered if it contained a predicted intrinsically disordered region with g40 consecutive amino acids and as putatively ordered otherwise. For notational convenience, we introduce a disorder operator d such that d(si) ) 1 if sequence si is putatively disordered, and d(si) ) 0 if it is putatively ordered. Relationship between Long Disorder Prediction and Protein Length. The likelihood of labeling a protein as putatively disordered increases with its length. To account for this length dependency, we estimated the probability, PL, that VL3E predicts a disordered region longer than 40 consecutive amino acids in a Swiss-Prot protein sequence of length L. Probability PL was determined by partitioning all Swiss-Prot proteins into groups based on their length. To reduce the effects of sequence redundancy, each sequence was weighted as the inverse of its family size; if sequence si was assigned to TribeMCL cluster c(si), we calculated ni as the total number of Swiss-Prot sequences assigned to this cluster and set its weight to w(si) ) 1/ni. In this manner, each cluster is given the same influence in estimation of PL, regardless of its size. To estimate PL, all Swiss-Prot sequences with length between L - l and L + l were grouped in set SL ) {si, L - l e |si| e L + l}. The probability PL was estimated as 1884

Journal of Proteome Research • Vol. 6, No. 5, 2007

PL )

∑ ∑

si∈SL

w(si)d(si)

si∈SLw(si)

Window size l allowed us to control the smoothness of PL function. In this study, we used window size equal to 20% of the sequence length, l ) 0.1L. We show the resulting curve in Figure 1 together with the same results when l ) 0. Extracting Disorder- and Order-Related Swiss-Prot Keywords. For each of the 710 Swiss-Prot keywords occurring in more than 20 Swiss-Prot proteins, we set to determine if it is enriched in putatively disordered or ordered proteins. For a keyword KWj, j ) 1, ..., 710, we first grouped all Swiss-Prot proteins annotated with the keyword to Sj. To take into consideration sequence redundancy, each sequence si ∈Sj was weighted based on the Swiss-Prot TribeMCL clusters. If sequence si was assigned to cluster c(si), we calculated nij as the total number of sequences from Sj that belonged to that cluster and set its weight to wj(i) ) 1/nij. Then, the fraction of putatively disordered proteins from Sj was calculated as Fj )

∑ ∑

si∈Sj

wj(si)d(si)

si∈SL

wj(si)

The question is how well this fraction fits the null model that is based on the length distribution PL. Let us define random variable Yj as Yj )

∑ ∑

si∈Sj

wj(si)X|si|

si∈SL

wj(si)

where XL is a Bernoulli random variable with P(XL ) 1) ) 1 P(XL ) 0) ) PL. In other words, Yj represents a distribution of fraction of putative disorder among randomly chosen SwissProt sequences with the same length distribution as those annotated with KWj. If Fj is in the left tail of the Yj distribution (i.e., the p-value P(Yj > Fj) is near 1), the keyword is enriched in ordered sequences, while if it is in the right tail (i.e., the p-value P(Yj > Fj) is near 0) it is enriched in disordered sequences. We denote all keywords with p-value < 0.05 as disorder-related and those with p-value > 0.95 as order-related. The distribution Yj is hard to derive analytically, so we randomly generated 1000 realizations and calculated the empirical p-value as the fraction of times these realizations were larger than Fj. We also calculated the mean µj and standard deviation σj of the 1000 realizations. We observed that, when |KWj| is large, distribution of Yj resembles a Gaussian distribution with mean µj and standard deviation σj. Using the Gaussian approximation, we calculated the Z-score of KWj as (Fj - µj)/σj and its p-value as 1/2(1 - erf(Zj/2)), where erf() is the error function. The Gaussian approximation is useful, since using the fraction of 1000 replicates is not accurate in estimating p-values below 0.01 or above 0.99. We report the Z-scores together with the empirical p-values in the results.

Results Estimating Correlation between Long Disordered Regions and Swiss-Prot Keywords. We applied the procedure described above to each of the 710 Swiss-Prot keywords occurring each in more than 20 Swiss-Prot proteins. These 710 keywords can be grouped into 11 functional categories, which are listed in

research articles


Figure 1. Fraction of putative disorder as a function of sequence length. The smoothed curve uses averaging window of size equal to 20% of the sequence length. Table 1. Summary of Association between the Prediction of Long Disordered Regions and Keywords for Each of the 11 Functional Categoriesa

functional category

Biological process Cellular component Coding sequence diversity Developmental stage Disease Domain Ligand Molecular function PTM Technical term Tissue TOTAL

no. no. no. keywords keywords keywords (P-value < 0.05) (P-value > 0.95)

301 77 9 4 17 34 72 143 37 12 4 710

174 23 0 0 0 9 41 37 11 7 0 302

73 33 6 3 11 21 17 51 18 2 3 238

a For each category, the table lists the total number of keywords associated with it (out of the 710 Swiss-Prot functional keywords), as well as number of keywords associated with predicted order and disorder.

Table 1. We denote keywords with p-value > 0.95 as disorderrelated and the ones with p-value < 0.05 as order-related. Keywords with p-value between 0.95 and 0.05 are ambiguous. These functions might depend on structured of disordered regions but simply exhibit signals that are too weak. Alternatively, these functions might depend on short regions of disorder or might require both ordered and disordered regions. The number of keywords strongly correlated with disorder and order is significantly larger than expected by the random model. This is evident by observing that, for a p-value threshold of 0.05, a random predictor would result in about 5% (∼36) of order-related and 5% of disorder-related keywords. These results suggest that presence or absence of disordered regions

is an important factor in the majority of biological functions and processes. Overall, this analysis shows that 238 Swiss-Prot functional keywords are disorder-related, whereas 302 are order-related. Interestingly, only two of the categories, “Biological Process” and “Ligand”, are enriched in order-related keywords, while the remaining 9 are enriched in the disorderrelated keywords. This result supports an earlier conjecture that disordered regions have a larger functional repertoire than the ordered regions.20 To further understand these function-disorder relationships, we carried out manual literature mining and studied a large number of individual experimental examples. To organize the presentation of these results, the keywords from various functional categories, which are most significantly associated with protein order and disorder, are arranged into specific groups (Tables 2-6). In each table, the disorder-function relationships are ranged by their Z-scores (see Materials and Methods). The Z-scores for all 710 functions are given in Supporting Information (see Table S1). One of the major goals here was to determine for each example whether the indicated function was carried out by regions of disorder or regions of structure. After all, the keyword-disorder correlations established by the method of Figure 2 do not determine whether the indicated association implies direct involvement of disorder with function or not. Biological Processes Associated with Intrinsically Disordered Proteins. The set of top 20 Swiss-Prot annotated cellular processes associated with predicted disorder are listed in Table 2. Presented below are several illustrative examples of biological processes from this list for which the associations with long disordered regions have been experimentally determined. The data below are organized in the following way: each discussed keyword is placed at the beginning of the corresponding Journal of Proteome Research • Vol. 6, No. 5, 2007 1885

research articles

Xie et al.

important component of the cellular environment, was shown to play a role in regulating differentiation and the differentiated phenotype of cells.40,41 An ECM is present within mammalian embryos from the two-cell stage and is a component of the environment of all cell types, although the composition of the ECM and the spatial relationships between cells and ECM differ between tissues. The ECM offers structural support for cells and can also act as a physical barrier or selective filter to soluble molecules.40 The ECM is composed of glycoproteins, glycosaminoglycans, and proteoglycans that are secreted and assembled locally into an organized network to which cells adhere.42 Cells interact with the ECM via numerous cell-binding sites located within individual ECM glycoproteins and ECM receptors. For example, in the case of fibronectin, the primary determinant of cell-binding activity for many cell types resides in the sequence GRGDSP, which occurs in one of the type III repeats that form the central domain of the molecule.43 Intriguingly, this cell-binding sequence exclusively consists of strongly disorder-promoting amino acid residues;37 thus, it very likely is intrinsically disordered. Importantly, it has been experimentally shown that the fibronectin binding domains from several different species of Gram-positive bacteria44 as well as the N-terminal domain of BBK32, a fibronectin-binding lipoprotein from Borrelia burgdorferi, the causative agent of Lyme disease, are all intrinsically disordered.45 Figure 2. Schematic representation of the algorithm for extracting disorder- and order-related keywords. Table 2. Top 20 Processes That Have Strongest Correlation with Predicted Disorder

keywords

Differentiation Transcription Transcription regulation Spermatogenesis DNA condensation Cell cycle mRNA processing mRNA splicing Mitosis Apoptosis Protein transport Meiosis Cell division Ubl conjugation pathway Wnt signaling pathway Neurogenesis Chromosome partition Ribosome biogenesis Chondrogenesis Growth regulation

number number average of of sequence proteins families length Z-score P-value

1406 11223 9758 332 317 4278 1575 716 718 810 3081 284 3466 1254 417 322 556 319 64 155

422 1653 1554 189 130 612 249 180 215 211 579 170 385 244 41 74 67 71 6 45

439.25 442.64 413.31 280.49 300.06 494.17 515.55 459.06 620.43 465.48 421.73 639.16 451.63 525.99 476.84 667.4 495.39 391.79 332.85 354.9

18.81 14.62 14.33 13.9 13.34 12.17 10.92 10.13 9.42 9.35 8.77 8.7 8.51 8.13 6.58 6.56 6.39 5.9 5.58 5.14

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

paragraph and italicized. If the following description involves other keywords discussed in this and the subsequent papers29,30 these keywords are presented using the italic font. Differentiation. In developmental biology, cellular differentiation describes the process by which different cell types are derived from a single fertilized egg cell. Differentiation is a continuously regulated process, with specific interactions between the cell and its environment playing a major role in maintaining stable expression of differentiation-specific genes.39 Obviously, numerous intracellular and extracellular proteins are involved in the differentiation control and regulation. For example, extracellular matrix (ECM), which is an 1886


Transcription and Transcription Regulation. Transcription, one of the key processes in the living cell through which a DNA sequence is enzymatically copied to produce a complementary RNA, is the transfer of genetic information from DNA into RNA. In the case of protein-encoding DNA, transcription initiates the process that ultimately leads to the translation of the genetic code into a functional protein. Transcription is strongly regulated by a number of proteins, especially transcription factors that include activators, repressors, and enhancer-binding factors. Transcription factors function through the recognition of specific DNA sequences and the recruitment and assembly of the transcription machinery. Thus, both protein-DNA and protein-protein recognition are central processes in the function of transcription factors. Several examples of intrinsically disordered proteins in transcriptional regulation have been reported.18,19 For example, the C-terminal activation domain of the bZIP proto-oncoprotein c-Fos is unstructured and highly mobile, yet this protein effectively suppresses transcription in vitro.46 The C-terminal domain of the transcriptional corepressor CtBP, which serves as a scaffold in the formation of a multiprotein complex hosting the essential components of both gene targeting and coordinated histone modifications, is also intrinsically disordered, as determined by using several complementary approaches (bioinformatics, NMR, CD, and smallangle X-ray scattering).47 Recent analysis of high-resolution structures of transcription factors in the Protein Data Bank revealed that these proteins are, on average, largely disordered molecules with over 60% of amino acids residing in ‘coiled’ configurations.48 The abundance of intrinsic disorder in transcriptional regulation was further demonstrated using a set of bioinformatics tools, including the Predictor Of Natural Disorder Regions (PONDR). This analysis showed that up to 94% of transcription factors have extended regions of intrinsic disorder. Furthermore, the analysis of the disorder distribution within the transcription factor data sets revealed that the degree of disorder is significantly higher in eukaryotic transcription factors than in it is in prokaryotic transcription factors.49 The complementary analysis of human transcriptional regulation


factors revealed that, although their average sequence is more than twice as long as that of prokaryotic proteins, the fraction of human sequences aligned to domains of known structure in PDB is less than half of that found for bacterial transcription factors,50 suggesting that the increased length of eukaryotic transcription factors results to a significant degree from the addition of disordered regions. Spermatogenesis. Spermatogenesis is the formation and development of mature spermatozoa from stem cells by meiosis and spermiogenesis. As spermatogenesis progresses, there is a widespread reorganization of the haploid genome followed by the extensive DNA condensation suggesting that the dynamic composition of chromatin is crucial for the activities of enzymes that act upon it. Histone variants such as H3.3, H2AX, and macroH2A play important roles at the various stages of spermiogenesis. Furthermore, post-translational modifications of different histones, including specifically modulated acetylation of histone H4 (acH4), ubiquitination of histones H2A and H2B (uH2A, uH2B), and phosphorylation of histone H3 (H3p), are also involved in the regulation of spermatogenesis.51 Furthermore, during the final stages of spermatogenesis, the DNA of sperm in most organisms is compacted due to the replacement of somatic-type histones by DNA-condensing sperm nuclear basic proteins (SNBPs), sperm histones (H type), protamine-like (PL type), and protamines (P type).52 Analysis of amino acid composition of PL-I sperm nuclear protein from Spisula solidissima revealed that it contains high amounts of lysine and arginine (24.8 and 23.1%, respectively).53 Also, the PL-I has been shown to possess a tripartite structure, consisting of N- and C-terminal flexible “tails” flanking a globular, trypsinresistant core of 75 amino acids.54-56 DNA Condensation. DNA condensation in vivo relies on electrostatics-driven interaction of DNA with small cations and/ or a number of abundant proteins including histones. In eukaryotes, the basic unit of chromatin (a condensed form of DNA) is commonly defined as a nucleosome, which is made up of DNA wrapped in two left-handed superhelical turns around a proteinaceous core.57 The nucleosome core contains eight histone proteins, two dimers of H2A-H2B that serve as molecular caps for the central (H3-H4)2 tetramer.58 Thus, nucleosome represents the first level of chromatin condensation and is often termed ‘beads on a string’. Other crucial components of chromatine are the linker (H1 family) histones, which bind to the DNA that enters and exits the nucleosome and which facilitate the shift in equilibrium of chromatin toward more condensed, higher order forms.57 It was established long ago that purified core histones, dissolved in water with no added salt, behave as polypeptides in an “extended loose form”.58-63 Recently, with the use of a combination of bioinformatics tools with several biophysical techniques, it has been shown that in low salt all bovine core histones are typical natively unfolded proteins; that is, they possess exceptionally high level of intrinsic disorder.64 Importantly, in the presence of high enough salt concentrations, core histones adopt a folded conformation.58-64 In the crystal structure, histones are highly helical proteins, with R-helices accounting for 65-70% of the total structure. Only 3% of residues can be assigned to short parallel β-sheets; the remainder, approximately 30%, is not ordered.65,66 It has been also emphasized that the Nterminal “tail” domains (NTDs)67 of the core histones and the C-terminal tail domain (CTD) of linker histones are intrinsically disordered; yet, they are able to bind to many different macromolecular partners in chromatin.68 Particularly, histone

research articles tails are known to be involved in the conformational changes of the nucleosome core particle (NCP) as well as in the structural phase transitions occurring at the supramolecular level. It is generally accepted that these tails interact with DNA at low salt and are extended outside of the particle at salt concentrations above ∼0.2 M monovalent salt.69 Analysis of the extension process of isolated NCP tails as a function of ionic strength has been reported. The addition of salt simultaneously screens Coulombic repulsive interactions between NCP and Coulombic attractive interactions between tails and DNA inside the NCP.70 Cell Cycle, Cell Division, Mitosis, Meiosis. The cell cycle consists of an ordered series of events between the two cell divisions and involve the growth, replication, and division of a eukaryotic cell. Depending on the type of cell, the cell division might result in two different outcomes: in the division of somatic cells (mitosis), daughter cells are identical to the parent cell and contain a complete copy of the parental chromosomes; in meiosis (the division of sex cells), the daughter cells contain a half of the genes of the parent. Progression through the cell cycle is controlled in part by the activity of cyclin-dependent kinases, which are considered to be the major timekeepers of cell division.71 Cdks are regulated by binding to their cyclin protein partners thus forming active heterodimeric complexes. Eight Cdk family members (Cdk1-Cdk8) and nine cyclins (AI) have been identified so far. Interestingly, each Cdk pairs with a separate cyclin class, most of which have at least two members.72,73 For example, Cdk1 together with cyclin B1 directs the G2/M transition. Exit from G1, in contrast, is primarily under the control of cyclin D/Cdk4/6. Finally, two other cyclins (A and E) that pair with Cdk2 are required for the G1/S transition and progression through the S phase.72,73 The activity of Cdks throughout the cell cycle is known to be precisely regulated by a combination of several mechanisms, including the control of cycle-dependent variations in the levels of activating partners, cyclins; coordination of Cdk phosphorylation and dephosphorylation; and variations in the levels of the Cdk inhibitor proteins, CKIs, which are responsible for the deactivation of the Cdk-cyclin complexes.71,74 Five major mammalian CKIs are known: p21Waf1/Cip1/Sdi1 and p27Kip1 inactivate Cdk2 and Cdk4 cyclin complexes by binding to them, p16INK4 and p15INK4B are specific for Cdk4 and Cdk6, whereas p57Kip2 is specific for Cdk2.71,74 The p21Waf1/Cip1/Sdi1,75 p27Kip1,76-78 and p57Kip2 CKIs79 are all intrinsically disordered proteins that undergo sequential folding upon binding to their functional partners. mRNA Processing and Splicing. An average gene in higher eukaryotes is very large due to the interruption of the coding sequence with large noncoding introns. Introns are known to be cotranscriptionally removed with great accuracy by premessenger RNA (mRNA) splicing. A large number of proteins are involved in generating specificity in pre-mRNA processing. Among the different pre-mRNA processing possibilities, alternative splicing is the most prevalent mechanism to generate proteomic diversity. The role of intrinsic disorder in alternative splicing is discussed in the second paper of this series.29 Astounding examples of extensively alternatively spliced genes include the Down syndrome cell adhesion molecule gene (Dscam) from Drosophila, the Neurexin and CD44 genes in humans, which can produce as many as about 38 000, 3000, and 1000 different splice forms, respectively.80-82 Splicing involves the stepwise assembly of five (U1, U2, U4, U5, and U6) small ribonucleoprotein particles (snRNPs) and a large Journal of Proteome Research • Vol. 6, No. 5, 2007 1887

research articles number of proteins onto the pre-mRNA to form a large complex called the spliceosome.83 The role of intrinsic disorder in the spliceosome function is discussed in the second paper of this series in the section entitled Cellular Components Associated with Intrinsically Disordered Proteins.29 Apoptosis. Apoptosis is the programmed death of a cell. Regulation and control of apoptosis is crucial for the normal functioning of the organism. On the other hand, cancer cells avoid apoptosis and continue to multiply in an unregulated manner. The tumor suppressor protein p53 represents an outstanding example of this concept. The p53 molecule regulates expression of genes involved in numerous cellular processes, including cell cycle progression, apoptosis induction, DNA repair, as well as others involved in responding to cellular stress.84 When p53 function is lost, either directly through mutation or indirectly through several other mechanisms, the cell often undergoes cancerous transformation.85,86 Cancers showing mutations in p53 are found in colon, lung, esophagus, breast, liver, brain, reticuloendothelial tissues, and hemopoietic tissues.85 When activated, p53 accumulates in the nucleus and binds to specific DNA sequences.86,87 It has been shown to induce or inhibit over 150 genes, including p21, GADD45, MDM2, IGFBP3, and BAX.87 The p53 protein can be divided into three functional domains: an amino-terminal transactivation region, a central DNA binding domain, and a carboxyterminal tetramerization and regulatory region. At the physiological temperature of 37 °C and in the absence of modifications or stabilizing partners, wild-type p53 is more than 50% unfolded.88 According to NMR analysis, the isolated transactivation domain lacks rigid structure,89,90 although it does possess an amphipathic helix that forms secondary structure part of the time, which can be stabilized by binding to Mdm291 or in the membrane environment.92 Besides Mdm2, the transactivation domain interacts with numerous other proteins including TFIID, TFIIH, RPA, CBP/p300, and CSN5/Jab1,84 thus, playing a crucial role in the regulation of p53 function. For example, p53 can be inhibited by interaction with E3 ubiquitin ligase Mdm2,93 which is bound to a short stretch of p53, specifically to residues 13-29.91 As this region of p53 is within the transactivation domain, p53 cannot activate or inhibit other genes when Mdm2 is bound. Thus, p53-Mdm2 interaction exemplifies a crucial mechanism of vital regulation via the binding-induced folding of one of the interacting partners. The C-terminal regulatory domain is also unstructured.94,95 The structures of the core domain bound to DNA96 and of several oncogenic mutants have been solved.97 The crystal and NMR structures of the tetramerization domain are also known.89,98,99 The high disorder content of p53 may help to explain its inherent instability and extreme oncogenic potential.100,101 The BH3-only proteins belong to the proapoptotic family of proteins that function as key initiators of the programmed cell death. The BH3-only members of the Bcl-2 family proteins (those with a single Bcl-2 homology (BH) domain), including Bim, Bid, Bmf, and Bad, are key initiators of apotosis, and they interact specifically with numerous binding partners.102 In a healthy cell, BH3-only molecules are either repressed or are present in an inactive state.103 The molecular mechanism of apoptosis initiation involves a stage of BH3-only proteins activation by a death stimulus, followed by their interaction and inactivation of prosurvival Bcl-2 proteins (such as CED-9 in Caenorhabdhitis elegans, and Bcl-xL in humans).102 Until recently, the structural knowledge about BH3-only proteins has been limited to peptide fragments bound to their targets. For 1888


Xie et al.

example, in the complex of Bcl-xL and a 33 residues long peptide corresponding to the BH3 domain of Bim, the peptide forms an R-helix upon binding to the hydrophobic groove of Bcl-xL.104 Analysis of crystal structures of BH3 domain peptides bound to the prosurvival proteins CED-9 and Bcl-xL revealed that the nature of this interaction is highly conserved despite only a low level of shared sequence identity with the conserved leucine and aspartic acid residues of the LXXXGDE motif, which defines BH3-only proteins, making critical contacts with conserved residues in the hydrophobic binding groove of CED-9 and Bcl-xL.104-107 However, in another structure, a 9-residue peptide of Bim forms a β-strand upon binding to the component of dynein motor complex, DLC1.108 Recently, with the use of methods such as CD, NMR, analytical centrifugation, size exclusion chromatography, and limited proteolysis, it has been established that the BH3-only proteins Bim, Bad, and Bmf are unstructured in the absence of binding partners.109 Intriguingly, the majority of the Bim residues remains disordered when this protein binds and inactivates prosurvival proteins, with only the short R-helical molecular recognition element110 becoming structured.109 Furthermore, detailed sequence analyses suggest that most BH3-only proteins are unstructured.109 The disorder of this proapoptotic protein family is likely to be important for several biological functions such as promiscuous binding, extensive splicing, and regulation via phosphorylation. Ubl Conjugation Pathway. Post-translational modification via the covalent attachment of ubiquitin and different ubiquitinlike proteins (Ubls, including SUMO, ISG15, Nedd8, and Atg8) is a crucial regulatory cellular mechanism, which plays a number of important roles in controlling cell division, signal transduction, embryonic development, endocytic trafficking, and the immune response.111 For example, conjugation of ubiquitin-like proteins (the Ubl conjugation pathway) to components of the transcriptional machinery is an important regulatory mechanism allowing switching between different activity states. While ubiquitination of transcription factors is associated with transcriptional activation, their SUMOylation is most often connected with transcriptional repression.112 Recent bioinformatic analysis of a limited number of known ubiquitination substrates showed that protein ubiquitination sites are preferentially located within surface-exposed, flexible loop regions.113 In addition, the sequence analysis of ubiquitination sites and regions adjacent to them showed that their properties such as low sequence complexity, high negative net charge, and low hydrophobicity are similar to those of intrinsically disordered regions (Iakoucheva and Radivojac, personal communication). An example of SUMOylation occurring within intrinsically unstructured region is the conjugation of transcription factor Ets-1 with SUMO-1.114 With the use of NMR spectroscopy, it has been shown that the SUMOylation motif of Ets-1 containing Lys15 is located within the unstructured N-terminal segment of Ets-1 preceding its PNT domain.114 The authors hypothesize that flexibility of the linking polypeptide sequence may be a general feature contributing to the recognition of SUMO-modified proteins by their downstream effectors.114 Wnt Signaling Pathway. Wnt is a critical pathway for embryogenesis, carcinogenesis, and cancer stem cells.115,116 Detailed information on this pathway can be found on the Wnt Homepage (http://www.stanford.edu/rnusse/wntwindow. html). The Wnt pathway shows evolutionary conservation across a wide range of species, ranging from the freshwater polyp Hydra to vertebrates.117 Mammals have 19 Wnt genes


that can be grouped into 12 subfamilies.118 Surprisingly, at least 11 of these subfamilies are present in Cnidaria (specifically, the sea anemone Nematostella vectensis) suggesting that they are not the result of any recent evolutionary diversification.119 This indicates that the acquisition of the Wnt subfamilies was an early development in the evolution of metazoa and likely occurred about 650 million years ago.119,120 The best-understood Wnt pathway is often called the Wnt/ β-catenin pathway, in which the Wnt signal leads to activation of the nuclear functions of β-catenin. These functions activate expression of a number of genes leading to cell survival, proliferation, or differentiation.121 A second vertebrate Wnt pathway, the Wnt/Ca2+ pathway, promotes intracellular Ca2+ release and regulates cell movements in development and in some cancers.122 A number of Wnt protein isoforms are generated by alternative splicing.123-125 Wnt proteins comprise a large family of highly conserved secreted growth factors that activate target-gene expression in both a short- and long-range manner and regulate cell-to-cell interactions during embryogenesis. Wnt signaling is involved in virtually every aspect of embryonic development and also controls homeostatic self-renewal in a number of adult tissues.115,117 Glycogen synthase kinase 3β (GSK3β) is a Ser/Thr protein kinase, which is one of the major players in the Wnt signaling pathway as GSK3β hyperphosphorylates β-catenin, thus, promoting its ubiquitination and targeted destruction.126 The crystal structure of human GSK3β (420 residues) has been solved at 2.8 Å.127 Clear electron density was only evident for the 351 residues from Lys35 to Ser386, and the segments of the polypeptide preceding Lys35 and following Ser386 were disordered in the crystal.127 The structure of the ordered part of GSK3β agrees with the consensus observed for “activationsegment” protein kinases, consisting of an N-terminal β-sheet domain, coupled with a C-terminal R-helical domain. The visible N-terminal domain (35-134) consists of a sevenstranded β-sheet, which folds to a closed orthogonal β-barrel. The core of the C-terminal R-helical domain (152-342) has a similar topology to the equivalent region in such mitogenactivated protein kinases, such as MAPK, as ERK2 and p38.127 It is important to emphasize that the major difference between the C-terminal R-helical domain of GSK3β and MAPK is the absence of the second helix in the hairpin segment from 276293 in the GSK3β domain. Furthermore, in GSK3β, this region represents a highly mobile and poorly defined 285-299 loop.127 Neurogenesis. Numerous proteins are involved in neurogenesis, the formation and development of nervous tissue. Among these proteins are the transcription factors Pax3,128 Pax6,129 Glis2,130 and Erm,131 which play an important regulatory role in this process. These transcription factors, like transcription factors in general, are highly disordered. For example, Pax3 has a highly flexible linker (53 amino acids) separating two DNA binding domains: a paired domain (128 amino acids) and a paired type homeodomain (60 amino acids).132 Similar to Pax3, transcription factor PAX6 has two DNA-binding domains, a paired domain, and a homeodomain (HD), joined by a glycinerich linker and followed by a proline-serine-threonine-rich (PST) transactivation region at the C-terminus.133 Structural analysis revealed that the central 250 amino acid residues of the transcription factor Erm has very little (if any) ordered structure.134 Chromosome Partition. Chromosome partition in two daughter cells is a complex process that involves a number of

research articles proteins. For example, proteins such as topoisomerase IV and XerCD recombinase, as well as MukB and FtsZ are related to chromosome partition in Escherichia coli.135 MukB exists as two thin rods (long antiparallel coiled coils) with globular domains at the ends emerging from the very flexible (read disordered) linker domain (123 amino acids).136 The flexibility of the hinge is crucial for the MukB function, as the arms can open up to 180°, separating the terminal domains by 100 nm, or close to near 0°, bringing the terminal globular domains together.136 Immune Response. The immune system is capable of generating specific antibodies against an almost infinite diversity of physiological or synthetic antigens. However, the repertoire of antibodies produced in any organism is fixed, suggesting that the immune system is an example of nearly unlimited functional multiplicity based on limited sequence diversity.137 The high flexibility of antigen-binding sites in the immunoglobulin, which provides the antibody with a unique capability to access a great variety of configurations with similar stabilities, was proposed to be the basis of this binding diversity.138 In more detail, the interplay between the intrinsic disorder, antigenic structure, and immunogenicity has been recently overviewed to emphasize the crucial role of intrinsic disorder in the development of immune response.22 For example, the conformational flexibility of antibodies drives their polyreactivity, thus, expanding the antigen-binding capacity of the antibody repertoire. On the other hand, short intrinsically disordered regions are likely important for the antigenicity of continuous determinants. Furthermore, the conformational flexibility and spatial adaptation play important roles in the antigen-antibody recognition and interaction. Finally, short intrinsically disordered regions are good antigens, whereas several long disordered regions and intrinsically disordered proteins initiate weak immune responses or are even completely nonimmunogenic.22 On the basis of these observations, it has been hypothesized that the role of intrinsic disorder in immunogenicity and antigenicity of a protein depends on the length of the disordered segment: short disordered regions (usually five to eight residues) are important for the development of the immune response to continuous epitopes, whereas long disordered regions (longer than 30 amino acids) are less likely to be immunogenic.22 The role of intrinsic disorder in autoimmune diseases has also been emphasized recently.139 The observation that a majority of the nuclear systemic autoantigens are extremely disordered proteins allowed the authors to introduce a new model of autoimmunity, disorder-based epitope spreading.139 Another example that illustrates the importance of disorder for immune response is the unstructured nature of the interferon tails.140 Finally, cytoplasmic domains of several immune receptors members of the family of multichain immune recognition receptors (MIRRs) (e.g., T-cell receptors (TCRs), B-cell receptors (BCRs), and the high-affinity IgE receptor) have signaling subunits carrying a so-called immunoreceptor tyrosine-based activation motif (ITAM).141-143 ITAM-containing cytoplasmic domains of signaling subunits of several MIRRs are intrinsically disordered.144,145 An intriguing feature of these signaling subunits is their tendency for the specific homooligomerization, which is not accompanied by their folding.145,146 Ribosome Biogenesis and rRNA Processing. Ribosomes are responsible for the production of the entire complement of proteins required for cellular maintenance, growth, and survival. Eukaryotic ribosomes contain four RNA molecules: 25S, Journal of Proteome Research • Vol. 6, No. 5, 2007 1889

research articles 18S, 5.8S, and 5S. The 5S rRNA is transcribed by RNA polymerase III, while the three other rRNA molecules are transcribed as a long 35S polycistronic precursor by RNA polymerase I.147 Ribosome biogenesis and rRNA processing are universal cellular processes, which encompass complicated series of events involving hundreds of transiently interacting components. It has been shown, for example, that in Saccharomyces cerevisiae the biogenesis of pre-18S ribosomal RNA is controlled by a large ribonucleoprotein (RNP) complex, which contains the U3 small nucleolar RNA (snoRNA) and 28 proteins.148 The analysis yielded five small subunit ribosomal proteins (Rps4, Rps6, Rps7, Rps14, and Rps28) among other proteins. Intriguingly, in eukaryotic cells, ribosomal protein S6 (Rps6) is the major phosphorylated protein on the small ribosomal subunit,149 suggesting that this protein might contain functionally important intrinsically disordered regions (see third paper in series, section dedicated to the post-translational modifications).30 Furthermore, bioinformatics analysis revealed that 14 of the U three proteins (Utps) bear different repeats comprising crucial regions of their protein-protein interaction domains (WD repeats, coiled-coil domains, HEAT repeats, and a crooked-neck-like (crn-like) tetratrico peptide repeat (TPR)).148 The crn-like TPR is found in several proteins involved in other RNA processing events including pre-mRNA splicing (Prp42, Prp6, and Clf1) and polyadenylation (RNA14).150 NMR analysis of the solution structure of the cytosolic TPR domain of Tom20 in the complex with the presequence peptide revealed that the C-terminal region of this protein (residues 105-145) is disordered.151 Chondrogenesis. This is the earliest phase of skeletal development, the process by which the cartilage is formed. Cartilage is an elastic connective tissue found in such parts of the body as the joints, outer ear, and larynx. Furthermore, cartilage represents the major constituent of the embryonic and young vertebrate skeleton, which is converted largely to bone with maturation. Chondrogenesis involves multiple steps, including mesenchymal cell recruitment and migration, condensation of progenitors, chondrocyte differentiation, and maturation, resulting in the formation of cartilage and bone during endochondral ossification.152 This complex process is precisely controlled by interactions with the surrounding matrix, growth, and differentiation factors, as well as other environmental factors responsible for the initiation or suppression of the cellular signaling pathways and for the regulation of transcription of specific genes.153 For example, the development of a vertebrate limb is controlled by the fibroblast growth factor, bone morphogenetic protein (BMP, a secreted signaling molecule, multifunctional growth factor belonging to the transforming growth factor β superfamily), Wnt, and hedgehog pathways.154 Recently, crucial roles of different mediators (including GADD45β, transcription factors of the Dlx, βHLH, leucine zipper, and AP-1 families, and the Wnt/β-catenin pathway) that interact at different stages during chondrogenesis have been revealed.153 Also, members of the mammalian RUNX protein family, which includes three transcription factors RUNX1, RUNX2, and RUNX3, are expressed during chondrogenesis. These transcription factors also play active roles in mesenchymal condensation, chondrocyte proliferation, and chondrocyte maturation and regulate transcription of target genes.155 Thus, regulation and control of chondrogenesis involve multiple players, many of which possess functional disordered regions. For example, the abundance and functional roles of intrinsic disorder in transcription factors were already 1890


Xie et al. Table 3. Top 20 Functions That Have Strongest Correlation with Predicted Disorder

keywords


Ribonucleoprotein 12236 Ribosomal protein 11692 Developmental protein 3260 Hormone 1187 Growth factor 785 Cytokine 899 Neuropeptide 268 Activator 3086 GAP protein 47 Antigen 1113 Repressor 2309 Chromatin regulator 334 Pyrogen 37 Vasoactive 125 Amphibian defense peptide 123 GTPase activation 311 Endorphin 42 Opioid peptide 24 Protein phosphatase 47 inhibitor Cyclin 182

412 330 721 161 84 110 209 573 2 455 449 100 2 39 148 70 4 4 8

150.55 140.58 477.93 141.13 255.7 213.28 95.08 428.47 232.96 437.48 374.46 801.24 262.59 160.39 50.64 831.03 226.68 216.96 366.51

22.13 20.63 19.28 15.58 11.16 10.21 9.65 9.04 7.42 6.99 6.92 6.7 6.44 5.56 5.44 5.36 5.35 5.14 5.07

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

25

430.58

4.88

1

discussed (see section entitled Transcription and transcription regulation), whereas the role of disorder in the leucine zippers will be discussed in the second paper of this series.29 Growth Regulation. Numerous proteins and pathways are implemented in growth regulation. For example, cyclin G was shown to be highly expressed in regenerating hepatocytes and motoneurons and in rapidly growing cancer cells and to have growth-promoting functions.156,157 Cyclin G interacts with cyclin-dependent kinase 5 (cdk5) and GAK, a cyclin G-associated kinase,158 as well as with with the B′ subclass of PP2A phosphatase.159 In addition, cyclin G directly interacts with Mdm2 and can stimulate the ability of PP2A to dephosphorylate Mdm2.159 Furthermore, cyclin G was one of the earliest p53 target genes to be identified.160 This suggests that cyclin G is a key regulator of the p53-Mdm2 network. The role of intrinsic disorder in p53 function was already discussed. While cyclins are structured, many are predicted to have long disordered fragments at the C-termini. Functions Associated with Intrinsically Disordered Proteins. Table 3 presents a list of the top 20 Swiss-Prot functional keywords associated with intrinsic disorder. Ribonucleoproteins. Numerous facts have been accumulated to demonstrate intrinsic disorder is crucial for function of different ribonucleoproteins. In fact, ribonucleoprotein assembly is nearly always accompanied by changes in the conformation of the interacting RNA, protein, or both.161-163 For example, the interdomain linkers of sex-lethal protein (SXL) possess significant disorder, which provide the RNA recognition motifs (RRMs) with a possibility to be flexibly tethered in solution.164 Another example is ribonuclease P (RNase P), a ribonucleoprotein complex containing one RNA subunit and at least one protein subunit. RNase P is involved in pre-tRNA processing.165 In E. coli, RNase P consists of a small (119 amino acid residue) C5 protein bound to the much larger (377 nucleotide) P RNA subunit.166 The C5 protein of E. coli is essentially disordered in buffer alone, but gains significant amount of ordered secondary structure in an anion-dependent manner.167 A similar behavior was also described for the Bacillus subtilis RNase P.168


Ribosomal Proteins. The assembly of the ribosome, which involves the sequential binding of numerous proteins via multiple pathways leading to large-scale changes in the conformation of the associated RNA and proteins, represents an extreme case involving dramatic structural changes induced by protein-RNA interaction.169-172 In fact, many ribosomal proteins have been shown to be significantly disordered prior binding to rRNA and to acquire ordered structure during ribosome formation.7,16,17 Developmental Proteins. R-Fetoprotein (AFP), a member of the family of albumin-like proteins, is a serum glycoprotein belonging to the intriguing class of onco-developmental polypeptides. AFP is homologous to human serum albumin (HSA).173 Similarly to HAS, AFP is able to bind a number of small molecules, including metal ions, estrogens, and different fatty acids (reviewed in ref 174). Importantly, the removal of all ligands from AFP is accompanied by a complete loss of rigid 3D structure (reviewed in ref 174). Hormones and Growth Factors. Growth hormone (GH), prolactin (PRL), and placental lactogen (PL), the pituitary hormones, are members of an extensive cytokine superfamily of hormones and receptors that share many of the same general structure-function relationships in expressing their biological activities.175 These hormones were shown to have two receptorbinding sites that have different topographies and electrostatic character. This feature is crucial for the regulation of these systems by producing binding surfaces with dramatically different binding affinities to the receptor extracellular domains. The receptor evidently possesses an exceptional conformational plasticity to be able to bind the topographically dissimilar sites on the hormone.175 Human parathyroid-hormone-related protein (hPTHrP) is a hormone that is overexpressed by a large number of tumors and is produced by a variety of normal cells. The N-terminal fragment (1-34) of hPTHrP is responsible for the major biological functions of this hormone. Furthermore, this fragment is mostly unstructured possessing only a small content of R-helical secondary structure.176 Secretin, a gut hormone consisting of 27 amino acid residues, was shown to be completely unfolded in aqueous solution but gain a fully ordered structure in the presence of 40% trifluoroethanol.177 Activators, Repressors, Cytokines, Protease Inhibitors, Antimicrobial Peptides and Amphibian Defense Peptides. Several other function-related keywords that are associated with intrinsically disordered proteins or regions are illustrated below. Protein AphA is a homodimeric member of a family of transcriptional activators. Transcriptional activators, including nuclear receptors, activate target genes through two broad classes of co-activators, those that remodel/modify chromatin and those that directly interfere with the general transcription machinery to facilitate formation and/or function of the preinitiation complexes. The AphA monomer is highly unstable by itself (i.e., likely it is highly disordered), and the dimer is formed in such a way that the two AphA chains wrap around each other,178 suggesting that the dimmer arose via a disorderto-order transition. The first 61 amino acid residues of the DNAfree lac repressor (i.e., a fragment which includes headpiece and the hinge region) are disordered and, thus, are unobserved in the crystal structure.179 Lymphotactin (Ltn) is unique member of family of pro-inflammatory activation-inducible cytokines. Ltn possesses a unique C-terminal extension, which is required for biological activity180,181 and which is disordered and highly mobile.182 Analysis of the Bowman-Birk protease inhibitor by Raman optical activity revealed that it possesses a “static”

research articles type of disorder similar to that in disordered states of poly(Llysine) and poly(L-glutamic acid).183 The pediocin-like class IIa bacteriocins, which are antimicrobial peptides from lactic acid bacteria,184,185 were shown to be significantly unstructured in water.186 Similarly, hylaseptin P1, an amphibian defense peptide, is in a random coil conformation in aqueous solutions.187 Neuropeptides. Pituitary adenylate cyclase activating polypeptide (PACAP),188 which occurs naturally in two forms consisting of a 38 amino acid peptide amide (PACAP38) and its 27 amino acid N-terminus (PACAP27), belongs to the secretin/glucagons/ vasoactive intestinal peptide (VIP) family.189 Structural analysis of PACAP38 and PACAP27 revealed that these two neuropeptides are mostly disordered and retain only small transitory amounts of stable structure in aqueous solution.190 Other opioid peptides are the enkephalins. The term enkephalin mainly refers to two peptides, [Met]-enkephalin and [Leu]-enkephalin, that both are products of the proenkephalin gene. [Met]enkephalin is Tyr-Gly-Gly-Phe-Met; [Leu]-enkephalin has Leu in place of Met. Recently performed structural characterization of methionine and leucine enkephalins by hydrogen/deuterium exchange and electrospray ionization tandem mass spectrometry revealed that the monomer forms of both peptides adopt an unfolded conformation in aqueous solvent, whereas they prefer β-turn secondary structure under the membranemimetic environment.191 GTPase Activation and GTPase-Activating Proteins (GAPs). The GTP-GDP conversion by guanine nucleotide binding proteins (GNBPs) represents an important timer in intracellular signaling and transport processes. GNBPs are highly abundant in different genomes. For example, there are at least 140 small GTPases encoded in human (including the Ras, Rho, Arf, Rab, and Ran GTPases), with various subclasses of this protein superfamily implicated in almost all aspects of cell biology, including proliferation, nucleocytoplasmic transport, differentiation, vesicle trafficking, cytoskeletal organization, and gene expression.192 These small GTPases are considered to be molecular switches, the cycling of which between active and inactive forms is regulated by cellular factors.192 There are two major classes of GNBP regulators, the guanine nucleotide exchange factors (GEFs), which promote the formation of active GTP-bound GTPases, and the GTPase activating proteins (GAPs), which promote GTPase inactivation by stimulating GTP-hydrolysis activity.193 In fact, the natural rate of GNBPmediated GTP hydrolysis is slow, but the reaction is accelerated by up to 5 orders of magnitude by the interaction of GNBPs with GAPs.194 At least 160 human genes have been recently predicted to encode proteins that resemble GAPs for various members of the Ras GPTase superfamily.195 Furthermore, ∼ 0.5% of all predicted human genes likely encode GAPs, suggesting that these proteins have widespread and important roles in GTPase regulation. Finally, such famous domains as ankyrin, BAR, BTK, CH, CNH, PDZ, PTB, RUN, SAM, SH2, SH3, WW, and many others are all GAPs.196 Many of the proteins in this group contain flexible linkers between their various domains. Chromatin Regulator. Several nuclear proteins serve as chromatin regulators, involved in modulation of chromosome structure, chromatin, and nucleosome remodeling and therefore playing a role in the controlling of gene transcription. Members of the HMGA family of non-histone chromatin proteins (formerly known as HMGI/Y proteins) serve as an illustrative example of such chromatin regulators.197 HMGA proteins are the founding members of a new class of regulatory Journal of Proteome Research • Vol. 6, No. 5, 2007 1891

research articles

Xie et al.

Table 4. Top 20 Processes That Have Strongest Correlation with Predicted Order keywords

number of proteins

number of families

average sequence length

Z-score

P-value

GMP biosynthesis Amino-acid biosynthesis Transport Electron transport Lipid A biosynthesis Aromatic hydrocarbons catabolism Glycolysis Purine biosynthesis Pyrimidine biosynthesis Carbohydrate metabolism Branched-chain amino acid biosynthesis Lipopolysaccharide biosynthesis Sugar transport Antibiotic resistance Lipid synthesis Tricarboxylic acid cycle Arginine biosynthesis Ion transport Rhamnose metabolism Peptidoglycan synthesis

225 7098 19888 4633 533 320 2255 1208 1310 1797 963 481 903 1203 2184 1013 1353 5275 85 1839

3 212 2199 346 13 105 50 28 27 180 26 102 109 177 122 54 17 459 4 38

473.11 361.5 378.13 272 291.25 300.36 390.64 445.46 383.27 404.2 404.12 335.93 387.37 354.24 328.02 460.88 414.06 464.46 372.84 372.73

-17.62 -17.11 -14.87 -13.72 -13.22 -12.37 -12.14 -11.89 -11.7 -11.68 -11.11 -11.09 -11 -10.66 -10.17 -10.04 -9.53 -9.37 -9.12 -9.03

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

elements called ‘architectural transcription factors’ that participate in a wide variety of cellular processes including regulation of inducible gene transcription, integration of retroviruses into chromosomes, the induction of neoplastic transformation, and promotion of metastatic progression of cancer cells.198 HMGA proteins are highly flexible and are characterized by the total lack of ordered structure.199-201 Pyrogen. Substances that can cause a rise in body temperature are known as pyrogens. Fever is the multiphasic response of elevation and decline of the body core temperature regulated by central thermoregulatory mechanisms localized in the preoptic area of the hypothalamus. Some cytokines (which are highly inducible, secreted proteins mediating intercellular communication in the nervous and immune system), including interleukin 1 (IL-1), interleukin 6 (IL-6), and the tumor necrosis factor alpha (TNFR), act as endogenous pyrogens.202 The role of intrinsic disorder in cytokine function was already discussed. Opioid Peptides and Endorphin. Opioid peptides are short natural peptides that mimic the effect of opiates in the brain and therefore are potent pain suppressants. Some opioid peptides (e.g., endorphin, dynorphin, and endorphin) are produced endogenously, some are produced by microbes (deltorphins and dermorphine), whereas others are absorbed from partially digested food (casomorphins, exorphins, and rubiscolins). Opioid peptides mediate their physiological and pharmacological effects through three major opioid receptor types (µ, δ, and κ),203 which are the members of the G proteincoupled receptor (GPCR) family.204 The 1H NMR spectra of human β-endorphin (the largest natural opioid peptide of 31 amino acid residues) indicate that the peptide exists in randomcoil conformation in aqueous solution but becomes helical in mixed solvent.205 Inhibitors. The activity of many important proteins is regulated by specific proteins inhibitors. It has been already mentioned that the functionality of Cdk inhibitor proteins relies on intrinsic disorder. Serine protease inhibitor elafin is a 57 amino acid residue peptide inhibiting human leukocyte elastase, porcine pancreatic elastase, and proteinase-3.206 Elafin was shown to be almost completely unfolded in aqueous solutions.207 Protein Phosphatase Inhibitors. Calcineurin (CaN) is a calcium- and calmodulin-dependent protein serine/threonine 1892


phosphate, which is critical for several important cellular processes, including T-cell activation.208 CaN is a heterodimer composed of subunits A and B (CaNA and CaN, respectively).208 CaN phosphatase activity is regulated by Ca2+ binding to CaNB and by Ca2+-induced binding of calmodulin (CaM) to CaNA.209 Activity of CaN is modulated by a number of factors, including an autoinhibitory domain (residues 457-479), which binds in the active site cleft in the absence of Ca2+/CaM and inhibits the enzyme, acting in concert with the CaM binding domain to confer CaM regulation.209 Analysis of CaN crystal structure revealed that CaNA residues 1-13, 374-468, and 487-521 are not visible in the electron density map, that is, disordered.209 Importantly, long disordered region 374-468 includes the CaMbinding helix.209 Therefore, this transient helix in CaN becomes bound and surrounded by CaM, turning on the CaN’s serine/ threonine phosphatase activity. Locating the CaN helix within the disordered region is essential for enabling CaM to surround its target upon binding.6 Cyclin. The progression of cells through the cell cycle is regulated by several specific proteins, known as cyclins and cyclin-dependent kinases. The concentrations of cyclins vary in a cyclical fashion during the cell cycle. They are produced or degraded as needed in order to drive the cell through the different stages of the cell cycle. The crucial role of intrinsic disorder in function and regulation of Cdk-cyclin complexes was already discussed. Biological Processes and Functions Associated with Ordered Proteins. Tables 4 and 5 list the top 20 biological processes and functions that are significantly associated with predicted order. An examination of order-correlated keywords suggests the presence of 5 major functional categories: (1) catalysis (this category includes all functions listed in Table 5; that is, oxidoreductase, transferase, hydrolase, lyase, glycosidase, kinase, isomerase, ligase, decarboxylase, glycosyltransferase, protease, acyltransferase, monooxygenase, aminotransferase, metalloprotease, methyltransferase, aminoacyl-tRNA synthetase, aminopeptidase, and dioxygenase); (2) transport (electron transport, sugar transport, and transport), (3) biosynthesis (aminoacid biosynthesis, GMP biosynthesis, gluconeogenesis, aminoacid biosynthesis, pyrimidine biosynthesis, peptidoglycan synthesis, lipopolysaccharide biosynthesis, aromatic amino acid biosynthesis, branched-chain amino acid biosynthesis, purine biosyn-

research articles

Functions of Intrinsic Disorder. Part 1 Table 5. Top 20 Functions That Have Strongest Correlation with Predicted Order

keywords


Oxidoreductase 14995 Transferase 26525 Lyase 7262 Hydrolase 20464 Isomerase 4487 Glycosidase 1826 Glycosyltransferase 2950 Acyltransferase 2239 Methyltransferase 3524 Kinase 7017 Ligase 8010 Decarboxylase 1293 Monooxygenase 1668 Metalloprotease 1100 Aminopeptidase 452 Dioxygenase 360 Aminoacyl-tRNA synthetase 3402 Protease 4423 Aminotransferase 955

992 1606 347 1995 220 244 261 179 224 322 230 63 73 109 39 66 37 380 28

376.63 445.17 377.92 430.68 383.98 444.73 437.53 402.83 349.6 448.29 529.41 345.26 444.87 553.73 509.17 433.2 571.83 549.7 420.27

-29.54 -24.25 -22.64 -21.75 -14.18 -13.98 -12.51 -10.85 -10.53 -10.22 -10.06 -9.66 -9.26 -7.89 -7.55 -7.32 -7.15 -7.1 -6.02

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

thesis, lipid a biosynthesis, and lipid synthesis); (4) metabolism (carbohydrate metabolism, tricarboxylic acid cycle, aromatic hydrocarbons catabolism, and one-carbon metabolism); (5) transmembrane proteins (porins). Note that catalysis overlaps strongly with biosynthesis and metabolism in that all of the proteins associated with these keywords are enzymes. The proteins associated with transport are often membrane proteins and are necessarily structured so that their backbone hydrogen bonds are formed in the low dielectric environment of the membrane. Other transport-associated proteins (which are not the membrane proteins) often need to bind very small molecules (or a single atom such as metals, or even electrons), which requires a precise coordination, and therefore, a welldefined structure. Proteins from the fifth category, porins, represent transmembrane proteins that are large enough to facilitate passive diffusion. They are prevalent in the outer membrane of the mitochondria and Gram-negative bacteria. Porins are almost entirely composed of β-sheets and control the diffusion of small metabolites like sugars, ions, and amino acids.210 They have been shown to have a highly stable structure using various characterization methods.211 For example, porin from Paracoccus denitrificans is extremely stable toward heat, pH, and chemical denaturants.212 This current result agrees with our previous observations that proteins involved in catalysis, transport, biosynthesis, and metabolism are less disordered than regulatory proteins. Interestingly, the smaller disorder content as well as the greater amount of structural information (e.g., a greater PDB coverage) has previously been reported for proteins from the first four functional categories as compared to highly disordered signaling and cancer-associated proteins.16 Thus, the current result agrees with our previous observations that proteins involved in catalysis, transport, biosynthesis, and metabolism are less disordered than regulatory proteins. Finally, one noticeable exception should be mentioned here. Although glycosidases are among the top 20 proteins with predicted functional order (Table 5), many of them in fact possess large disordered regions, even though their catalytic function requires a well-defined structure. This is especially true for cellulases (Biological process: cellulose degradation, strong correlation with predicted order, see Table S1 Supporting

Table 6. All (11) Swiss-Prot Keywords Associated with at Least One of the 98 Confirmed Long Disordered Protein Regions8,9 a

function

number of regions associate with function in literatures8,9

Z-ratio in Swiss-Prot database

Protein-DNA interaction Phosphorylation Structural mortar Ubiquitination Protein-rRNA interaction Fatty acylation Protein-genomic RNA binding Glycosylation Methylation ADP-ribosylation Protein-tRNA interaction

19 16 >10 7 5 4 3 3 1 1 1

18.2 27.1 3.0 8.7 11.5 6.0 17.7 6.8 2.9 2.0 1.8

a For each function, number of the associated regions (out of 98) and Z-ratio are listed.

Information) for which protein disorder has been experimentally determined.213,214 These cellulases are composed of a catalytic domain, linked to a cellulose binding domain through a long disordered linker (109 amino acid residues in Cel5G, an endoglucanase from Pseudoalteromonas haloplanktis), which could be considered as an entropic spring. In fact, the SAXS analysis of dimensions, shape, and conformation of Cel5G fulllength in solution and especially of the linker between the catalytic module and the cellulose-binding module revealed that the linker is unstructured, and unusually long and flexible.213 This modular organization and the presence of a disordered linker are crucial to optimize the biphasic process of crystalline cellulose degradation. Another example of an enzyme that possesses functional disordered regions is retinaldehyde dehydrogenase II (RalDH2).215 This enzyme converts retinal to the transcriptional regulator retinoic acid in the developing embryo. It has been shown that a 20-amino acid span in the substrate access channel is disordered, but folds during the course of catalysis and provides a means for an enzyme that requires a large substrate access channel to restrict access to the catalytic machinery by smaller compounds that might potentially enter the active site and be metabolized.215 Therefore, RalDH2 represents a unique example of a protein that exhibits a catalytic activity in which a large disordered region folds upon catalysis. Comparison of the Identified Disorder Functions with Literature Findings. Recently, literature analysis identified 28 functions associated with 98 confirmed disordered regions containing 30 or longer contiguous disorder residues.8,9 These functions were grouped into four broad categories: molecular recognition, molecular assembly, protein modification, and entropic chains. Entropic chains carry out functions that depend directly on the disordered state, and so such functions are simply outside the capabilities of fully folded structures.8,9 The use of partially folded subunits for molecular assembly appears to have significant advantages compared with the use of ordered subunits.21,22 Molecular recognition appears to be a common function for both ordered and disordered proteins: molecular recognition by disordered proteins may be primarily used for signaling, whereas recognition by ordered proteins may be primarily used for catalysis,8,9 or for the assembly of functional complexes. Finally, sites of some types of posttranslational modification frequently occur within the regions with very strong preference for disorder.8-11,18,19,22,216 Journal of Proteome Research • Vol. 6, No. 5, 2007 1893

research articles Out of 28 functions associated with the confirmed disordered regions, 13 were also found in the Swiss-Prot keyword list. Eleven of these functions are strongly correlated with predicted disorder with p-value above 0.95. Table 6 lists these 11 functions together with their Z-scores. Furthermore, previously reported16 strong functional correlations to intrinsic disorder were also found using the method proposed in this study. These results strongly support the validity of the proposed statistical methodology for finding disorder-correlated functions.

Conclusions We proposed a statistical approach that estimates the correlations between protein structure and protein function. As an application, we studied the relationship between intrinsic protein disorder and function in the Swiss-Prot database. Overall, 238 Swiss-Prot functional keywords were discovered to be strongly associated with predicted intrinsic disorder, and 302 function keywords were shown to be strongly correlated with predicted order. We validated a significant fraction of these findings by comparing them to literature data. The numerous correlations between the known experimental data and the results of the analysis demonstrate the general validity of our approach. However, a more thorough comparison with literature data will be needed to determine the frequencies with which exceptions occur as compared to the observed trends. In our original, manual curation of disorder-function correlations, we determined that each function was unambiguously associated with a specific region of disorder.8,9 While the current methodology does not determine whether each function is directly associated with a disordered segment, the literature data verify that these functions are indeed, for the most part, associated with regions of disorder. Of special interest is that disease-related proteins were shown to have the high correlation with disordered regions of proteins (see the last paper of this series30). Overall, this approach provides an innovative and relevant method to examine protein structure-function relationships.

Acknowledgment. The authors express their deepest gratitude to Celeste Brown and Predrag Radivojac for numerous valuable discussions. This work was supported by grants from the National Institutes of Health LM007688-0A1 (A.K.D. and Z.O.) and GM071714-01A2 (A.K.D and V.N.U.), and by the Indiana Genomics Initiative (INGEN) (A.K.D.). INGEN is supported in part by the Lilly Endowment, Inc. The Programs of the Russian Academy of Sciences for the “Molecular and cellular biology” and “Fundamental science for medicine” provided partial support to V.N.U., and L.M.I. was supported by the NSF grant MCB 0444818. Supporting Information Available: Table S1 represents the Z-scores for all 710 functions. Red color indicates strong correlation with predicted disorder; yellow color indicates strong correlation with predicted order. This material is available free of charge via the Internet at http://pubs.acs.org. References (1) Fischer, E. Einfluss der Configuration auf die Wirkung der Enzyme. Ber. Dtsch. Chem. Ges. 1894, 27, 2985-2993. (2) Wu, H. Studies on denaturation of proteins XIIIsa theory of denaturation. Chin. J. Physiol. 1931, 1, 219-234. (3) Mirsky, A. E.; Pauling, L. On the structure of native, denatured, and coagulated proteins. Proc. Natl. Acad. Sci. U.S.A. 1936, 22, 439-447.

1894


Xie et al. (4) Sela, M.; White, F. H., Jr.; Anfinsen, C. B. Reductive cleavage of disulfide bridges in ribonuclease. Science 1957, 125, 691-692. (5) Berman, H. M.; Westbrook, J.; Feng, Z.; Gilliland, G.; Bhat, T. N.; Weissig, H.; Shindyalov, I. N.; Bourne, P. E. The Protein Data Bank. Nucleic Acids Res. 2000, 28, 235-242. (6) Dunker, A. K.; Lawson, J. D.; Brown, C. J.; Williams, R. M.; Romero, P.; Oh, J. S.; Oldfield, C. J.; Campen, A. M.; Ratliff, C. M.; Hipps, K. W.; Ausio, J.; Nissen, M. S.; Reeves, R.; Kang, C.; Kissinger, C. R.; Bailey, R. W.; Griswold, M. D.; Chiu, W.; Garner, E. C.; Obradovic, Z. Intrinsically disordered protein. J. Mol. Graphics Modell. 2001, 19, 26-59. (7) Uversky, V. N. Natively unfolded proteins: a point where biology waits for physics. Protein Sci. 2002, 11, 739-756. (8) Dunker, A. K.; Brown, C. J.; Lawson, J. D.; Iakoucheva, L. M.; Obradovic, Z. Intrinsic disorder and protein function. Biochemistry 2002, 41, 6573-6582. (9) Dunker, A. K.; Brown, C. J.; Obradovic, Z. Identification and functions of usefully disordered proteins. Adv. Protein Chem. 2002, 62, 25-49. (10) Tompa, P. Intrinsically unstructured proteins. Trends Biochem. Sci. 2002, 27, 527-533. (11) Wright, P. E.; Dyson, H. J. Intrinsically unstructured proteins: reassessing the protein structure-function paradigm. J. Mol. Biol. 1999, 293, 321-331. (12) Dunker, A. K.; Obradovic, Z. The protein trinity-linking function and disorder. Nat. Biotechnol. 2001, 19, 805-806. (13) Uversky, V. N. What does it mean to be natively unfolded? Eur. J. Biochem. 2002, 269, 2-12. (14) Oldfield, C. J.; Cheng, Y.; Cortese, M. S.; Brown, C. J.; Uversky, V. N.; Dunker, A. K. Comparing and combining predictors of mostly disordered proteins. Biochemistry 2005, 44, 1989-2000. (15) Dunker, A. K.; Obradovic, Z.; Romero, P.; Garner, E. C.; Brown, C. J. Intrinsic protein disorder in complete genomes. Genome Inf. Ser. 2000, 11, 161-171. (16) Iakoucheva, L. M.; Brown, C. J.; Lawson, J. D.; Obradovic, Z.; Dunker, A. K. Intrinsic disorder in cell-signaling and cancerassociated proteins. J. Mol. Biol. 2002, 323, 573-584. (17) Uversky, V. N.; Gillespie, J. R.; Fink, A. L. Why are “natively unfolded” proteins unstructured under physiologic conditions? Proteins 2000, 41, 415-427. (18) Dyson, H. J.; Wright, P. E. Coupling of folding and binding for unstructured proteins. Curr. Opin. Struct. Biol. 2002, 12, 54-60. (19) Dyson, H. J.; Wright, P. E. Intrinsically unstructured proteins and their functions. Nat. Rev. Mol. Cell Biol. 2005, 6, 197-208. (20) Daughdrill, G. W.; Pielak, G. J.; Uversky, V. N.; Cortese, M. S.; Dunker, A. K. Natively disordered proteins. In Handbook of Protein Folding; Buchner, J., Kiefhaber, T., Eds.; Wiley-VCH, Verlag GmbH & Co. KGaA: Weinheim, Germany, 2005; pp 271353. (21) Dunker, A. K.; Cortese, M. S.; Romero, P.; Iakoucheva, L. M.; Uversky, V. N. Flexible nets. The roles of intrinsic disorder in protein interaction networks. FEBS J. 2005, 272, 5129-5148. (22) Uversky, V. N.; Oldfield, C. J.; Dunker, A. K. Showing your ID: intrinsic disorder as an ID for recognition, regulation and cell signaling. J. Mol. Recognit. 2005, 18, 343-384. (23) Haynes, C.; Oldfield, C. J.; Ji, F.; Klitgord, N.; Cusick, M. E.; Radivojac, P.; Uversky, V. N.; Vidal, M.; Iakoucheva, L. M. Intrinsic disorder is a common feature of hub proteins from four eukaryotic interactomes. PLoS Comput. Biol. 2006, 2, e100. (24) Vucetic, S.; Obradovic, Z.; Vacic, V.; Radivojac, P.; Peng, K.; Iakoucheva, L. M.; Cortese, M. S.; Lawson, J. D.; Brown, C. J.; Sikes, J. G.; Newton, C. D.; Dunker, A. K. DisProt: a database of protein disorder. Bioinformatics 2005, 21, 137-140. (25) Dyson, H. J.; Wright, P. E. According to current textbooks, a welldefined three-dimensional structure is a prerequisite for the function of a protein. Is this correct? IUBMB Life 2006, 58, 107109. (26) Peng, K.; Vucetic, S.; Radivojac, P.; Brown, C. J.; Dunker, A. K.; Obradovic, Z. Optimizing long intrinsic disorder predictors with protein evolutionary information. J. Bioinf. Comput. Biol. 2005, 3, 35-60. (27) Boeckmann, B.; Bairoch, A.; Apweiler, R.; Blatter, M. C.; Estreicher, A.; Gasteiger, E.; Martin, M. J.; Michoud, K.; O’Donovan, C.; Phan, I.; Pilbout, S.; Schneider, M. The SWISS-PROT protein knowledgebase and its supplement TrEMBL in 2003. Nucleic Acids Res. 2003, 31, 365-370. (28) Ward, J. J.; Sodhi, J. S.; McGuffin, L. J.; Buxton, B. F.; Jones, D. T. Prediction and functional analysis of native disorder in proteins from the three kingdoms of life. J. Mol. Biol. 2004, 337, 635-645.

Functions of Intrinsic Disorder. Part 1 (29) Xie, H.; Vucetic, S.; Iakoucheva, L. M.; Oldfield, C. J.; Dunker, A. K.; Obradovic, Z.; Uversky, V. N. Functional anthology of intrinsic disorder. 2. Cellular components, domains, technical terms, developmental processes, and coding sequence diversity associated with long disordered regions. J. Proteome Res. 2007, 5, 18991916. (30) Xie, H.; Vucetic, S.; Iakoucheva, L. M.; Oldfield, C. J.; Dunker, A. K.; Obradovic, Z.; Uversky, V. N. Functional anthology of intrinsic disorder. 3. Ligands, postranslational modifications, and diseases associated with long disordered regions. J. Proteome Res. 2007, 5, 1917-1932. (31) O’Donovan, C.; Martin, M. J.; Glemet, E.; Codani, J. J.; Apweiler, R. Removing redundancy in SWISS-PROT and TrEMBL. Bioinformatics 1999, 15, 258-259. (32) Enright, A. J.; Van, Dongen, S.; Ouzounis, C. A. An efficient algorithm for large-scale detection of protein families. Nucleic Acids Res. 2002, 30, 1575-1584. (33) Altschul, S. F.; Gish, W.; Miller, W.; Myers, E. W.; Lipman, D. J. Basic local alignment search tool. J. Mol. Biol. 1990, 215, 403410. (34) Romero, P.; Obradovic, Z.; Dunker, A. K. Sequence data analysis for long disordered regions prediction in the calcineurin family. Genome Inf. Ser. 1997, 8, 110-124. (35) Romero, P.; Obradovic, Z.; Kissinger, C.; Villafranca, J. E.; Dunker, A. K. Identifying disordered regions in proteins from amino acid sequence. Proc. Int. Conf. Neural Networks 1997, 1, 90-95. (36) Xie, Q.; Arnold, G. E.; Romero, P.; Obradovic, Z.; Garner, E.; Dunker, A. K. The sequence attribute method for determining relationships between sequence and protein disorder. Genome Inf. Ser. 1998, 9, 193-200. (37) Romero, P.; Obradovic, Z.; Li, X.; Garner, E. C.; Brown, C. J.; Dunker, A. K. Sequence complexity of disordered protein. Proteins 2001, 42, 38-48. (38) Sickmeier, M.; Hamilton, J. A.; LeGall, T.; Vacic, V.; Cortese, M. S.; Tantos, A.; Szabo, B.; Tompa, P.; Chen, J.; Uversky, V. N.; Obradovic, Z.; Dunker, A. K. DisProt: the Database of Disordered Proteins. Nucleic Acids Res. 2007, 35, D786-793. (39) Blau, H. M.; Baltimore, D. Differentiation requires continuous regulation. J. Cell Biol. 1991, 112, 781-783. (40) Adams, J. C.; Watt, F. M. Regulation of development and differentiation by the extracellular matrix. Development 1993, 117, 1183-1198. (41) Watt, F. M. The extracellular matrix and cell shape. Trends Biochem. Sci. 1986, 11, 482-485. (42) Hay, E. D. Cell Biology of Extracellular Matrix; Plenum Press: New York, 1981. (43) Ruoslahti, E.; Pierschbacher, M. D. New perspectives in cell adhesion: RGD and integrins. Science 1987, 238, 491-497. (44) House-Pompeo, K.; Xu, Y.; Joh, D.; Speziale, P.; Hook, M. Conformational changes in the fibronectin binding MSCRAMMs are induced by ligand binding. J. Biol. Chem. 1996, 271, 13791384. (45) Kim, J. H.; Singvall, J.; Schwarz-Linek, U.; Johnson, B. J.; Potts, J. R.; Hook, M. BBK32, a fibronectin binding MSCRAMM from Borrelia burgdorferi, contains a disordered region that undergoes a conformational change on ligand binding. J. Biol. Chem. 2004, 279, 41706-41714. (46) Campbell, K. M.; Terrell, A. R.; Laybourn, P. J.; Lumb, K. J. Intrinsic structural disorder of the C-terminal activation domain from the bZIP transcription factor Fos. Biochemistry 2000, 39, 2708-2713. (47) Haynes, C.; Iakoucheva, L. M. Serine/arginine-rich splicing factors belong to a class of intrinsically disordered proteins. Nucleic Acids Res. 2006, 34, 305-312. (48) Bhalla, J.; Storchan, G. B.; Maccarthy, C. M.; Uversky, V. N.; Tcherkasskaya, O. Local flexibility in molecular function paradigm. Mol. Cell. Proteomics 2006, 5, 1212-1223. (49) Liu, J.; Perumal, N. B.; Oldfield, C. J.; Su, E. W.; Uversky, V. N.; Dunker, A. K. Intrinsic disorder in transcription factors. Biochemistry 2006, 45, 6873-6888. (50) Minezaki, Y.; Homma, K.; Kinjo, A. R.; Nishikawa, K. Human transcription factors contain a high fraction of intrinsically disordered regions essential for transcriptional regulation. J. Mol. Biol. 2006, 359, 1137-1149. (51) Lewis, J. D.; Abbott, D. W.; Ausio, J. A haploid affair: core histone transitions during spermatogenesis. Biochem. Cell Biol. 2003, 81, 131-140. (52) Ausio, J. Histone H1 and the evolution of the nuclear spermspecific proteins. In Advances in Spermatozoal Phylogeny and Taxonomy; Jamieson, B. G. M., Ausio, J., Justine, J. L., Eds.; Memoires du Museum National d’Histoire Naturelle: Paris, France, 1995.

research articles (53) Ausio, J.; Subirana, J. A. A high molecular weight nuclear basic protein from the bivalve mollusc Spisula solidissima. J. Biol. Chem. 1982, 257, 2802-2805. (54) Ausio, J.; Toumadje, A.; McParland, R.; Becker, R. R.; Johnson, W. C. Jr.; van Holde, K. E. Structural characterization of the trypsin-resistant core in the nuclear sperm-specific protein from Spisula solidissima. Biochemistry 1987, 26, 975-982. (55) Lewis, J. D.; Ausio, J. Protamine-like proteins: evidence for a novel chromatin structure. Biochem. Cell Biol. 2002, 80, 353-361. (56) Lewis, J. D.; McParland, R.; Ausio, J. PL-I of Spisula solidissima, a highly elongated sperm-specific histone H1. Biochemistry 2004, 43, 7766-7775. (57) Harvey, A. C.; Downs, J. A. What functions do linker histones provide? Mol. Microbiol. 2004, 53, 771-775. (58) Isenberg, I. Histones. Annu. Rev. Biochem. 1979, 48, 159-191. (59) Boublik, M.; Bradbury, E. M.; Crane-Robinson, C. An investigation of the conformational changes in histones F1 and F2a1 by proton magnetic resonance spectroscopy. Eur. J. Biochem. 1970, 14, 486497. (60) Li, H. J.; Wickett, R.; Craig, A. M.; Isenberg, I. Conformational changes in histone IV. Biopolymers 1972, 11, 375-397. (61) Wickett, R. R.; Li, H. J.; Isenberg, I. Salt effects on histone IV conformation. Biochemistry 1972, 11, 2952-2957. (62) D’Anna, J. A. Jr.; Isenberg, I. Conformational changes of histone ARE(F3, III). Biochemistry 1974, 13, 4987-4992. (63) D’Anna, J. A., Jr.; Isenberg, I. Conformational changes of histone LAK (f2a2). Biochemistry 1974, 13, 2093-2098. (64) Munishkina, L. A.; Fink, A. L.; Uversky, V. N. Conformational prerequisites for formation of amyloid fibrils from histones. J. Mol. Biol. 2004, 342, 1305-1324. (65) Luger, K.; Mader, A. W.; Richmond, R. K.; Sargent, D. F.; Richmond, T. J. Crystal structure of the nucleosome core particle at 2.8 A resolution. Nature 1997, 389, 251-260. (66) Arents, G.; Burlingame, R. W.; Wang, B. C.; Love, W. E.; Moudrianakis, E. N. The nucleosomal core histone octamer at 3.1 Å resolution: a tripartite protein assembly and a left-handed superhelix. Proc. Natl. Acad. Sci. U.S.A. 1991, 88, 10148-10152. (67) Hansen, J. C. Conformational dynamics of the chromatin fiber in solution: determinants, mechanisms, and functions. Annu. Rev. Biophys. Biomol. Struct. 2002, 31, 361-392. (68) Hansen, J. C.; Lu, X.; Ross, E. D.; Woody, R. W. Intrinsic protein disorder, amino acid composition, and histone terminal domains. J. Biol. Chem. 2006, 281, 1853-1856. (69) Walker, I. O. Differential dissociation of histone tails from core chromatin. Biochemistry 1984, 23, 5622-5628. (70) Mangenot, S.; Leforestier, A.; Vachette, P.; Durand, D.; Livolant, F. Salt-induced conformation and interaction changes of nucleosome core particles. Biophys. J. 2002, 82, 345-356. (71) Morgan, D. O. Principles of CDK regulation. Nature 1995, 374, 131-134. (72) Morgan, D. O. Cyclin-dependent kinases: engines, clocks, and microprocessors. Annu. Rev. Cell Dev. Biol. 1997, 13, 261-291. (73) Nigg, E. A. Mitotic kinases as regulators of cell division and its checkpoints. Nat. Rev. Mol. Cell Biol. 2001, 2, 21-32. (74) Pavletich, N. P. Mechanisms of cyclin-dependent kinase regulation: structures of Cdks, their cyclin activators, and Cip and INK4 inhibitors. J. Mol. Biol. 1999, 287, 821-828. (75) Kriwacki, R. W.; Hengst, L.; Tennant, L.; Reed, S. I.; Wright, P. E. Structural studies of p21Waf1/Cip1/Sdi1 in the free and Cdk2bound state: conformational disorder mediates binding diversity. Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 11504-11509. (76) Flaugh, S. L.; Lumb, K. J. Effects of macromolecular crowding on the intrinsically disordered proteins c-Fos and p27(Kip1). Biomacromolecules 2001, 2, 538-540. (77) Bienkiewicz, E. A.; Adkins, J. N.; Lumb, K. J. Functional consequences of preorganized helical structure in the intrinsically disordered cell-cycle inhibitor p27(Kip1). Biochemistry 2002, 41, 752-759. (78) Lacy, E. R.; Filippov, I.; Lewis, W. S.; Otieno, S.; Xiao, L.; Weiss, S.; Hengst, L.; Kriwacki, R. W. p27 binds cyclin-CDK complexes through a sequential mechanism involving binding-induced protein folding. Nat. Struct. Mol. Biol. 2004, 11, 358-364. (79) Adkins, J. N.; Lumb, K. J. Intrinsic structural disorder and sequence features of the cell cycle inhibitor p57Kip2. Proteins 2002, 46, 1-7. (80) Schmucker, D.; Clemens, J. C.; Shu, H.; Worby, C. A.; Xiao, J.; Muda, M.; Dixon, J. E.; Zipursky, S. L. Drosophila Dscam is an axon guidance receptor exhibiting extraordinary molecular diversity. Cell 2000, 101, 671-684.

Journal of Proteome Research • Vol. 6, No. 5, 2007 1895

research articles (81) Zhu, J.; Shendure, J.; Mitra, R. D.; Church, G. M. Single molecule profiling of alternative pre-mRNA splicing. Science 2003, 301, 836-838. (82) Tabuchi, K.; Sudhof, T. C. Structure and evolution of neurexin genes: insight into the mechanism of alternative splicing. Genomics 2002, 79, 849-859. (83) Jurica, M. S.; Moore, M. J. Pre-mRNA splicing: awash in a sea of proteins. Mol. Cell 2003, 12, 5-14. (84) Anderson, C. W.; Appella, E. Signaling to the p53 tumor suppressor through pathways activated by genotoxic and nongenotoxic stress. In Handbook of Cell Signaling; Bradshaw, R. A., Dennis, E. A., Eds.; Academic Press: New York, 2004; pp 237-247. (85) Hollstein, M.; Sidransky, D.; Vogelstein, B.; Harris, C. C. p53 mutations in human cancers. Science 1991, 253, 49-53. (86) Balint, E. E.; Vousden, K. H. Activation and activities of the p53 tumour suppressor protein. Br. J. Cancer 2001, 85, 1813-1823. (87) Zhao, R.; Gish, K.; Murphy, M.; Yin, Y.; Notterman, D.; Hoffman, W. H.; Tom, E.; Mack, D. H.; Levine, A. J. Analysis of p53-regulated gene expression patterns using oligonucleotide arrays. Genes Dev. 2000, 14, 981-993. (88) Bell, S.; Klein, C.; Muller, L.; Hansen, S.; Buchner, J. p53 contains large unstructured regions in its native state. J. Mol. Biol. 2002, 322, 917-927. (89) Lee, H.; Mok, K. H.; Muhandiram, R.; Park, K. H.; Suk, J. E.; Kim, D. H.; Chang, J.; Sung, Y. C.; Choi, K. Y.; Han, K. H. Local structural elements in the mostly unstructured transcriptional activation domain of human p53. J. Biol. Chem. 2000, 275, 29426-29432. (90) Dawson, R.; Muller, L.; Dehner, A.; Klein, C.; Kessler, H.; Buchner, J. The N-terminal domain of p53 is natively unfolded. J. Mol. Biol. 2003, 332, 1131-1141. (91) Kussie, P. H.; Gorina, S.; Marechal, V.; Elenbaas, B.; Moreau, J.; Levine, A. J.; Pavletich, N. P. Structure of the MDM2 oncoprotein bound to the p53 tumor suppressor transactivation domain. Science 1996, 274, 948-953. (92) Rosal, R.; Pincus, M. R.; Brandt-Rauf, P. W.; Fine, R. L.; Michl, J.; Wang, H. NMR solution structure of a peptide from the mdm-2 binding domain of the p53 protein that is selectively cytotoxic to cancer cells. Biochemistry 2004, 43, 1854-1861. (93) Vargas, D. A.; Takahashi, S.; Ronai, Z. Mdm2: A regulator of cell growth and death. Adv. Cancer Res. 2003, 89, 1-34. (94) Ayed, A.; Mulder, F. A.; Yi, G. S.; Lu, Y.; Kay, L. E.; Arrowsmith, C. H. Latent and active p53 are identical in conformation. Nat. Struct. Biol. 2001, 8, 756-760. (95) Weinberg, R. L.; Freund, S. M.; Veprintsev, D. B.; Bycroft, M.; Fersht, A. R. Regulation of DNA binding of p53 by its C-terminal domain. J. Mol. Biol. 2004, 342, 801-811. (96) Cho, Y.; Gorina, S.; Jeffrey, P. D.; Pavletich, N. P. Crystal structure of a p53 tumor suppressor-DNA complex: understanding tumorigenic mutations. Science 1994, 265, 346-355. (97) Joerger, A. C.; Ang, H. C.; Veprintsev, D. B.; Blair, C. M.; Fersht, A. R. Structures of p53 cancer mutants and mechanism of rescue by second-site suppressor mutations. J. Biol. Chem. 2005, 280, 16030-16037. (98) Clore, G. M.; Ernst, J.; Clubb, R.; Omichinski, J. G.; Kennedy, W. M.; Sakaguchi, K.; Appella, E.; Gronenborn, A. M. Refined solution structure of the oligomerization domain of the tumour suppressor p53. Nat. Struct. Biol. 1995, 2 321-333. (99) Lee, W.; Harvey, T. S.; Yin, Y.; Yau, P.; Litchfield, D.; Arrowsmith, C. H. Solution structure of the tetrameric minimum transforming domain of p53. Nat. Struct. Biol. 1994, 1, 877-890. (100) Canadillas, J. M.; Tidow, H.; Freund, S. M.; Rutherford, T. J.; Ang, H. C.; Fersht, A. R. Solution structure of p53 core domain: structural basis for its instability. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 2109-2114. (101) Veprintsev, D. B.; Freund, S. M.; Andreeva, A.; Rutledge, S. E.; Tidow, H.; Canadillas, J. M.; Blair, C. M.; Fersht, A. R. Core domain interactions in full-length p53 in solution. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 2115-2119. (102) Hinds, M. G.; Day, C. L. Regulation of apoptosis: uncovering the binding determinants. Curr. Opin. Struct. Biol. 2005, 15, 690699. (103) Puthalakath, H.; Strasser, A. Keeping killers on a tight leash: transcriptional and post-translational control of the pro-apoptotic activity of BH3-only proteins. Cell Death Differ. 2002, 9, 505512. (104) Liu, X.; Dai, S.; Zhu, Y.; Marrack, P.; Kappler, J. W. The structure of a Bcl-xL/Bim fragment complex: implications for Bim function. Immunity 2003, 19, 341-352.

1896


Xie et al. (105) Sattler, M.; Liang, H.; Nettesheim, D.; Meadows, R. P.; Harlan, J. E.; Eberstadt, M.; Yoon, H. S.; Shuker, S. B.; Chang, B. S.; Minn, A. J.; Thompson, C. B.; Fesik, S. W. Structure of Bcl-xL-Bak peptide complex: recognition between regulators of apoptosis. Science 1997, 275, 983-986. (106) Petros, A. M.; Nettesheim, D. G.; Wang, Y.; Olejniczak, E. T.; Meadows, R. P.; Mack, J.; Swift, K.; Matayoshi, E. D.; Zhang, H.; Thompson, C. B.; Fesik, S. W. Rationale for Bcl-xL/Bad peptide complex formation from structure, mutagenesis, and biophysical studies. Protein Sci. 2000, 9, 2528-2534. (107) Yan, N.; Gu, L.; Kokel, D.; Chai, J.; Li, W.; Han, A.; Chen, L.; Xue, D.; Shi, Y. Structural, biochemical, and functional analyses of CED-9 recognition by the proapoptotic proteins EGL-1 and CED4. Mol. Cell 2004, 15, 999-1006. (108) Fan, J.; Zhang, Q.; Tochio, H.; Li, M.; Zhang, M. Structural basis of diverse sequence-dependent target recognition by the 8 kDa dynein light chain. J. Mol. Biol. 2001, 306, 97-108. (109) Hinds, M. G.; Smits, C.; Fredericks-Short, R.; Risk, J. M.; Bailey, M.; Huang, D. C.; Day, C. L. Bim, Bad and Bmf: intrinsically unstructured BH3-only proteins that undergo a localized conformational change upon binding to prosurvival Bcl-2 targets. Cell Death Differ. 2006, 14, 128-136. (110) Oldfield, C. J.; Cheng, Y.; Cortese, M. S.; Romero, P.; Uversky, V. N.; Dunker, A. K. Coupled folding and binding with R-helixforming molecular recognition elements. Biochemistry 2005, 44, 12454-12470. (111) Huang, D. T.; Walden, H.; Duda, D.; Schulman, B. A. Ubiquitinlike protein activation. Oncogene 2004, 23, 1958-1971. (112) Hay, R. T. Role of ubiquitin-like proteins in transcriptional regulation. Ernst Schering Res. Found. Workshop 2006, 173-192. (113) Catic, A.; Collins, C.; Church, G. M.; Ploegh, H. L. Preferred in vivo ubiquitination sites. Bioinformatics 2004, 20, 3302-3307. (114) Macauley, M. S.; Errington, W. J.; Scharpf, M.; Mackereth, C. D.; Blaszczak, A. G.; Graves, B. J.; McIntosh, L. P. Beads-on-a-string, characterization of ETS-1 sumoylated within its flexible Nterminal sequence. J. Biol. Chem. 2006, 281, 4164-4172. (115) Logan, C. Y.; Nusse, R. The Wnt signaling pathway in development and disease. Annu. Rev. Cell Dev. Biol. 2004, 20, 781-810. (116) Kelleher, F. C.; Fennelly, D.; Rafferty, M. Common critical pathways in embryogenesis and cancer. Acta Oncol. 2006, 45, 375-388. (117) Clevers, H. Wnt/beta-catenin signaling in development and disease. Cell 2006, 127, 469-480. (118) Prud’homme, B.; Lartillot, N.; Balavoine, G.; Adoutte, A.; Vervoort, M. Phylogenetic analysis of the Wnt gene family. Insights from lophotrochozoan members. Curr. Biol. 2002, 12, 1395. (119) Kusserow, A.; Pang, K.; Sturm, C.; Hrouda, M.; Lentfer, J.; Schmidt, H. A.; Technau, U.; von Haeseler, A.; Hobmayer, B.; Martindale, M. Q.; Holstein, T. W. Unexpected complexity of the Wnt gene family in a sea anemone. Nature 2005, 433, 156-160. (120) Gordon, M. D.; Nusse, R. Wnt signaling: multiple pathways, multiple receptors, and multiple transcription factors. J. Biol. Chem. 2006, 281, 22429-22433. (121) Moon, R. T.; Bowerman, B.; Boutros, M.; Perrimon, N. The promise and perils of Wnt signaling through beta-catenin. Science 2002, 296, 1644-1646. (122) Kohn, A. D.; Moon, R. T. Wnt and calcium signaling: betacatenin-independent pathways. Cell Calcium 2005, 38, 439-446. (123) Katoh, M.; Kirikoshi, H.; Saitoh, T.; Sagara, N.; Koike, J. Alternative splicing of the WNT-2B/WNT-13 gene. Biochem. Biophys. Res. Commun. 2000, 275, 209-216. (124) Pospisil, H.; Herrmann, A.; Butherus, K.; Pirson, S.; Reich, J. G.; Kemmner, W. Verification of predicted alternatively spliced Wnt genes reveals two new splice variants (CTNNB1 and LRP5) and altered Axin-1 expression during tumour progression. BMC Genomics 2006, 7, 148. (125) Struewing, I. T.; Toborek, A.; Mao, C. D. Mitochondrial and nuclear forms of Wnt13 are generated via alternative promoters, alternative RNA splicing, and alternative translation start sites. J. Biol. Chem. 2006, 281, 7282-7293. (126) Ding, Y.; Dale, T. Wnt signal transduction: kinase cogs in a nanomachine? Trends Biochem. Sci. 2002, 27, 327-329. (127) Dajani, R.; Fraser, E.; Roe, S. M.; Young, N.; Good, V.; Dale, T. C.; Pearl, L. H. Crystal structure of glycogen synthase kinase 3 beta: structural basis for phosphate-primed substrate specificity and autoinhibition. Cell 2001, 105, 721-732. (128) Apuzzo, S.; Gros, P. The paired domain of pax3 contains a putative homeodomain interaction pocket defined by cysteine scanning mutagenesis. Biochemistry 2006, 45, 7154-7161.

Functions of Intrinsic Disorder. Part 1 (129) Haubst, N.; Berger, J.; Radjendirane, V.; Graw, J.; Favor, J.; Saunders, G. F.; Stoykova, A.; Gotz, M. Molecular dissection of Pax6 function: the specific roles of the paired domain and homeodomain in brain development. Development 2004, 131, 6131-6140. (130) Zhang, F.; Nakanishi, G.; Kurebayashi, S.; Yoshino, K.; Perantoni, A.; Kim, Y. S.; Jetten, A. M. Characterization of Glis2, a novel gene encoding a Gli-related, Kruppel-like transcription factor with transactivation and repressor functions. Roles in kidney development and neurogenesis. J. Biol. Chem. 2002, 277, 10139-10149. (131) Paratore, C.; Brugnoli, G.; Lee, H. Y.; Suter, U.; Sommer, L. The role of the Ets domain transcription factor Erm in modulating differentiation of neural crest stem cells. Dev. Biol. 2002, 250, 168-180. (132) Chalepakis, G.; Wijnholds, J.; Gruss, P. Pax-3-DNA interaction: flexibility in the DNA binding and induction of DNA conformational changes by paired domains. Nucleic Acids Res. 1994, 22, 3131-3137. (133) Mishra, R.; Gorlov, I. P.; Chao, L. Y.; Singh, S.; Saunders, G. F. PAX6, paired domain influences sequence recognition by the homeodomain. J. Biol. Chem. 2002, 277, 49488-49494. (134) Mauen, S.; Huvent, I.; Raussens, V.; Demonte, D.; Baert, J. L.; Tricot, C.; Ruysschaert, J. M.; Van Lint, C.; Moguilevsky, N.; de Launoit, Y. Expression, purification, and structural prediction of the Ets transcription factor ERM. Biochim. Biophys. Acta 2006, 1760, 1192-1201. (135) Hiraga, S. Chromosome partition in Escherichia coli. Curr. Opin. Genet. Dev. 1993, 3, 789-801. (136) Melby, T. E.; Ciampaglio, C. N.; Briscoe, G.; Erickson, H. P. The symmetrical structure of structural maintenance of chromosomes (SMC) and MukB proteins: long, antiparallel coiled coils, folded at a flexible hinge. J. Cell Biol. 1998, 142, 1595-1604. (137) James, L. C.; Tawfik, D. S. Conformational diversity and protein evolution-a 60-year-old hypothesis revisited. Trends Biochem. Sci. 2003, 28, 361-368. (138) Pauling, L. A theory of the structure and process of formation of antibodies. J. Am. Chem. Soc. 1940, 62, 2643-2657. (139) Carl, P. L.; Temple, B. R.; Cohen, P. L. Most nuclear systemic autoantigens are extremely disordered proteins: implications for the etiology of systemic autoimmunity. Arthritis Res. Ther. 2005, 7, R1360-1374. (140) Slutzki, M.; Jaitin, D. A.; Yehezkel, T. B.; Schreiber, G. Variations in the unstructured c-terminal tail of interferons contribute to differential receptor binding and biological activity. J. Mol. Biol. 2006, 360, 1019-1030. (141) Sigalov, A. B. Multichain immune recognition receptor signaling: different players, same game? Trends Immunol. 2004, 25, 583-589. (142) Sigalov, A. B. Immune cell signaling: a novel mechanistic model reveals new therapeutic targets. Trends Pharmacol. Sci. 2006, 27, 518-524. (143) Sigalov, A. Multi-chain immune recognition receptors: spatial organization and signal transduction. Semin. Immunol. 2005, 17, 51-64. (144) Sigalov, A. B.; Aivazian, D. A.; Uversky, V. N.; Stern, L. J. Lipidbinding activity of intrinsically unstructured cytoplasmic domains of multichain immune recognition receptor signaling subunits. Biochemistry 2006, 45, 15731-15739. (145) Sigalov, A.; Aivazian, D.; Stern, L. Homooligomerization of the cytoplasmic domain of the T cell receptor zeta chain and of other proteins containing the immunoreceptor tyrosine-based activation motif. Biochemistry 2004, 43, 2049-2061. (146) Sigalov, A. B.; Zhuravleva, A. V.; Orekhov, V. Y. Binding of intrinsically disordered proteins is not necessarily accompanied by a structural transition to a folded form. Biochimie 2006, in press. (147) Wehner, K. A.; Baserga, S. J. The sigma(70)-like motif: a eukaryotic RNA binding domain unique to a superfamily of proteins required for ribosome biogenesis. Mol. Cell 2002, 9, 329-339. (148) Dragon, F.; Gallagher, J. E.; Compagnone-Post, P. A.; Mitchell, B. M.; Porwancher, K. A.; Wehner, K. A.; Wormsley, S.; Settlage, R. E.; Shabanowitz, J.; Osheim, Y.; Beyer, A. L.; Hunt, D. F.; Baserga, S. J. A large nucleolar U3 ribonucleoprotein required for 18S ribosomal RNA biogenesis. Nature 2002, 417, 967-970. (149) Hernandez, V. P.; Higgins, L.; Schwientek, M. S.; Fallon, A. M. The histone-like C-terminal extension in ribosomal protein S6 in Aedes and Anopheles mosquitoes is encoded within the distal portion of exon 3. Insect Biochem. Mol. Biol. 2003, 33, 901-910.

research articles (150) Chung, S.; McLean, M. R.; Rymond, B. C. Yeast ortholog of the Drosophila crooked neck protein promotes spliceosome assembly through stable U4/U6.U5 snRNP addition. RNA 1999, 5, 10421054. (151) Abe, Y.; Shodai, T.; Muto, T.; Mihara, K.; Torii, H.; Nishikawa, S.; Endo, T.; Kohda, D. Structural basis of presequence recognition by the mitochondrial protein import receptor Tom20. Cell 2000, 100, 551-560. (152) Olsen, B. R.; Reginato, A. M.; Wang, W. Bone development. Annu. Rev. Cell Dev. Biol. 2000, 16, 191-220. (153) Goldring, M. B.; Tsuchimochi, K.; Ijiri, K. The control of chondrogenesis. J. Cell. Biochem. 2006, 97, 33-44. (154) Tickle, C. Molecular basis of vertebrate limb patterning. Am. J. Med. Genet. 2002, 112, 250-255. (155) Yoshida, C. A.; Komori, T. Role of Runx proteins in chondrogenesis. Crit. Rev. Eukaryotic Gene Expression 2005, 15, 243-254. (156) Morita, N.; Kiryu, S.; Kiyama, H. p53-independent cyclin G expression in a group of mature neurons and its enhanced expression during nerve regeneration. J. Neurosci. 1996, 16, 59615966. (157) Skotzko, M.; Wu, L.; Anderson, W. F.; Gordon, E. M.; Hall, F. L. Retroviral vector-mediated gene transfer of antisense cyclin G1 (CYCG1) inhibits proliferation of human osteogenic sarcoma cells. Cancer Res. 1995, 55, 5493-5498. (158) Kanaoka, Y.; Kimura, S. H.; Okazaki, I.; Ikeda, M.; Nojima, H. GAK: a cyclin G associated kinase contains a tensin/auxilin-like domain. FEBS Lett. 1997, 402, 73-80. (159) Okamoto, K.; Kamibayashi, C.; Serrano, M.; Prives, C.; Mumby, M. C.; Beach, D. p53-dependent association between cyclin G and the B’ subunit of protein phosphatase 2A. Mol. Cell. Biol. 1996, 16, 6593-6602. (160) Okamoto, K.; Beach, D. Cyclin G is a transcriptional target of the p53 tumor suppressor protein. EMBO J. 1994, 13, 4816-4822. (161) Williamson, J. R. Induced fit in RNA-protein recognition. Nat. Struct. Biol. 2000, 7, 834-837. (162) Leulliot, N.; Varani, G. Current topics in RNA-protein recognition: control of specificity and biological function through induced fit and conformational capture. Biochemistry 2001, 40, 7947-7956. (163) Chen, Y.; Varani, G. Protein families and RNA recognition. Febs J. 2005, 272, 2088-2097. (164) Crowder, S. M.; Kanaar, R.; Rio, D. C.; Alber, T. Absence of interdomain contacts in the crystal structure of the RNA recognition motifs of Sex-lethal. Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 4892-4897. (165) Torres-Larios, A.; Swinger, K. K.; Pan, T.; Mondragon, A. Structure of ribonuclease Psa universal ribozyme. Curr. Opin. Struct. Biol. 2006, 16, 327-335. (166) Harris, M. E.; Christian, E. L. Recent insights into the structure and function of the ribonucleoprotein enzyme ribonuclease P. Curr. Opin. Struct. Biol. 2003, 13, 325-333. (167) Guo, X.; Campbell, F. E.; Sun, L.; Christian, E. L.; Anderson, V. E.; Harris, M. E. RNA-dependent folding and stabilization of C5 protein during assembly of the E. coli RNase P holoenzyme. J. Mol. Biol. 2006, 360, 190-203. (168) Henkels, C. H.; Kurz, J. C.; Fierke, C. A.; Oas, T. G. Linked folding and anion binding of the Bacillus subtilis ribonuclease P protein. Biochemistry 2001, 40, 2777-2789. (169) Nissen, P.; Hansen, J.; Ban, N.; Moore, P. B.; Steitz, T. A. The structural basis of ribosome activity in peptide bond synthesis. Science 2000, 289, 920-930. (170) Recht, M. I.; Williamson, J. R. RNA tertiary structure and cooperative assembly of a large ribonucleoprotein complex. J. Mol. Biol. 2004, 344, 395-407. (171) Schroeder, R.; Barta, A.; Semrad, K. Strategies for RNA folding and assembly. Nat. Rev. Mol. Cell Biol. 2004, 5, 908-919. (172) Klein, D. J.; Moore, P. B.; Steitz, T. A. The roles of ribosomal proteins in the structure assembly, and evolution of the large ribosomal subunit. J. Mol. Biol. 2004, 340, 141-177. (173) Deutsch, H. F. Chemistry and biology of alpha-fetoprotein. Adv. Cancer Res. 1991, 56, 253-312. (174) Gillespie, J. R.; Uversky, V. N. Structure and function of alphafetoprotein: a biophysical overview. Biochim. Biophys. Acta 2000, 1480, 41-56. (175) Kossiakoff, A. A. The structural basis for biological signaling, regulation, and specificity in the growth hormone-prolactin system of hormones and receptors. Adv. Protein Chem. 2004, 68, 147-169.

Journal of Proteome Research • Vol. 6, No. 5, 2007 1897

research articles (176) Gronwald, W.; Schomburg, D.; Tegge, W.; Wray, V. Assessment by 1H NMR spectroscopy of the structural behaviour of human parathyroid-hormone-related protein(1-34) and its close relationship with the N-terminal fragments of human parathyroid hormone in solution. Biol. Chem. 1997, 378, 1501-1508. (177) Gronenborn, A. M.; Bovermann, G.; Clore, G. M. A 1H-NMR study of the solution conformation of secretin. Resonance assignment and secondary structure. FEBS Lett. 1987, 215, 88-94. (178) De, Silva, R. S.; Kovacikova, G.; Lin, W.; Taylor, R. K.; Skorupski, K.; Kull, F. J. Crystal structure of the virulence gene activator AphA from Vibrio cholerae reveals it is a novel member of the winged helix transcription factor superfamily. J. Biol. Chem. 2005, 280, 13779-13783. (179) Lewis, M.; Chang, G.; Horton, N. C.; Kercher, M. A.; Pace, H. C.; Schumacher, M. A.; Brennan, R. G.; Lu, P. Crystal structure of the lactose operon repressor and its complexes with DNA and inducer. Science 1996, 271, 1247-1254. (180) Hedrick, J. A.; Zlotnik, A. Lymphotactin: a new class of chemokine. Methods Enzymol. 1997, 287, 206-215. (181) Marcaurelle, L. A.; Mizoue, L. S.; Wilken, J.; Oldham, L.; Kent, S. B.; Handel, T. M.; Bertozzi, C. R. Chemical synthesis of lymphotactin: a glycosylated chemokine with a C-terminal mucin-like domain. Chemistry 2001, 7, 1129-1132. (182) Kuloglu, E. S.; McCaslin, D. R.; Markley, J. L.; Volkman, B. F. Structural rearrangement of human lymphotactin, a C chemokine, under physiological solution conditions. J. Biol. Chem. 2002, 277, 17863-17870. (183) Smyth, E.; Syme, C. D.; Blanch, E. W.; Hecht, L.; Vasak, M.; Barron, L. D. Solution structure of native proteins with irregular folds from Raman optical activity. Biopolymers 2001, 58, 138-151. (184) Ennahar, S.; Sashihara, T.; Sonomoto, K.; Ishizaki, A. Class IIa bacteriocins: biosynthesis, structure and activity. FEMS Microbiol. Rev. 2000, 24, 85-106. (185) Eijsink, V. G.; Axelsson, L.; Diep, D. B.; Havarstein, L. S.; Holo, H.; Nes, I. F. Production of class II bacteriocins by lactic acid bacteria; an example of biological warfare and communication. Antonie Van Leeuwenhoek 2002, 81, 639-654. (186) Kaur, K.; Andrew, L. C.; Wishart, D. S.; Vederas, J. C. Dynamic relationships among type IIa bacteriocins: temperature effects on antimicrobial activity and on structure of the C-terminal amphipathic R helix as a receptor-binding region. Biochemistry 2004, 43, 9009-9020. (187) Prates, M. V.; Sforca, M. L.; Regis, W. C.; Leite, J. R.; Silva, L. P.; Pertinhez, T. A.; Araujo, A. L.; Azevedo, R. B.; Spisni, A.; Bloch, C., Jr. The NMR-derived solution structure of a new cationic antimicrobial peptide from the skin secretion of the anuran Hyla punctata. J. Biol. Chem. 2004, 279, 13018-13026. (188) Miyata, A.; Arimura, A.; Dahl, R. R.; Minamino, N.; Uehara, A.; Jiang, L.; Culler, M. D.; Coy, D. H. Isolation of a novel 38 residuehypothalamic polypeptide which stimulates adenylate cyclase in pituitary cells. Biochem. Biophys. Res. Commun. 1989, 164, 567574. (189) Arimura, A. Pituitary adenylate cyclase activating polypeptide (PACAP): discovery and current status of research. Regul. Pept. 1992, 37, 287-303. (190) Wray, V.; Kakoschke, C.; Nokihara, K.; Naruse, S. Solution structure of pituitary adenylate cyclase activating polypeptide by nuclear magnetic resonance spectroscopy. Biochemistry 1993, 32, 5832-5841. (191) Cai, X.; Dass, C. Structural characterization of methionine and leucine enkephalins by hydrogen/deuterium exchange and electrospray ionization tandem mass spectrometry. Rapid Commun. Mass Spectrom. 2005, 19, 1-8. (192) Takai, Y.; Sasaki, T.; Matozaki, T. Small GTP-binding proteins. Physiol. Rev. 2001, 81 153-208. (193) Symons, M.; Settleman, J. Rho family GTPases: more than simple switches. Trends Cell Biol. 2000, 10, 415-419. (194) Scheffzek, K.; Ahmadian, M. R. GTPase activating proteins: structural and functional insights 18 years after discovery. Cell Mol. Life Sci. 2005, 62, 3014-3038. (195) Bernards, A. GAPs galore! A survey of putative Ras superfamily GTPase activating proteins in man and Drosophila. Biochim. Biophys. Acta 2003, 1603, 47-82. (196) Bernards, A.; Settleman, J. GAP control: regulating the regulators of small GTPases. Trends Cell Biol. 2004, 14, 377-385.

1898


Xie et al. (197) Grosschedl, R.; Giese, K.; Pagel, J. HMG domain proteins: architectural elements in the assembly of nucleoprotein structures. Trends Genet. 1994, 10, 94-100. (198) Reeves, R.; Beckerbauer, L. HMGI/Y proteins: flexible regulators of transcription and chromatin structure. Biochim. Biophys. Acta 2001, 1519, 13-29. (199) Lehn, D. A.; Elton, T. S.; Johnson, K. R.; Reeves, R. A conformational study of the sequence specific binding of HMG-I (Y) with the bovine interleukin-2 cDNA. Biochem. Int. 1988, 16, 963971. (200) Evans, J. N.; Nissen, M. S.; R. R. Assignment of the 1H NMR spectrum of a consensus DNA-binding peptide from the HMG-I protein. Bull. Magn. Reson. 1992, 14, 171-174. (201) Evans, J. N.; Zajicek, J.; Nissen, M. S.; Munske, G.; Smith, V.; Reeves, R. 1H and 13C NMR assignments and molecular modelling of a minor groove DNA-binding peptide from the HMG-I protein. Int. J. Pept. Protein Res. 1995, 45, 554-560. (202) Conti, B.; Tabarean, I.; Andrei, C.; Bartfai, T. Cytokines and fever. Front. Biosci. 2004, 9, 1433-1449. (203) Janecka, A.; Kruszynski, R. Conformationally restricted peptides as tools in opioid receptor studies. Curr. Med. Chem. 2005, 12, 471-481. (204) Chaturvedi, K.; Christoffers, K. H.; Singh, K.; Howells, R. D. Structure and regulation of opioid receptors. Biopolymers 2000, 55, 334-346. (205) Lichtarge, O.; Jardetzky, O.; Li, C. H. Secondary structure determination of human β-endorphin by 1H NMR spectroscopy. Biochemistry 1987, 26, 5916-5925. (206) Tsunemi, M.; Kato, H.; Nishiuchi, Y.; Kumagaye, S.; Sakakibara, S. Synthesis and structure-activity relationships of elafin, an elastase-specific inhibitor. Biochem. Biophys. Res. Commun. 1992, 185, 967-973. (207) Francart, C.; Dauchez, M.; Alix, A. J.; Lippens, G. Solution structure of R-elafin, a specific inhibitor of elastase. J. Mol. Biol. 1997, 268, 666-677. (208) Rusnak, F.; Mertz, P. Calcineurin: form and function. Physiol. Rev. 2000, 80, 1483-1521. (209) Kissinger, C. R.; Parge, H. E.; Knighton, D. R.; Lewis, C. T.; Pelletier, L. A.; Tempczyk, A.; Kalish, V. J.; Tucker, K. D.; Showalter, R. E.; Moomaw, E. W.; et al. Crystal structures of human calcineurin and the human FKBP12-FK506-calcineurin complex. Nature 1995, 378, 641-644. (210) Delcour, A. H. Structure and function of pore-forming betabarrels from bacteria. J. Mol. Microbiol. Biotechnol. 2002, 4, 1-10. (211) Sukumaran, S.; Hauser, K.; Maier, E.; Benz, R.; Mantele, W. Structure-function correlation of outer membrane protein porin from Paracoccus denitrificans. Biopolymers 2006, 82, 344-348. (212) Sukumaran, S.; Hauser, K.; Maier, E.; Benz, R.; Mantele, W. Tracking the unfolding and refolding pathways of outer membrane protein porin from Paracoccus denitrificans. Biochemistry 2006, 45, 3972-3980. (213) Violot, S.; Aghajari, N.; Czjzek, M.; Feller, G.; Sonan, G. K.; Gouet, P.; Gerday, C.; Haser, R.; Receveur-Brechot, V. Structure of a full length psychrophilic cellulase from Pseudoalteromonas haloplanktis revealed by X-ray diffraction and small angle X-ray scattering. J. Mol. Biol. 2005, 348, 1211-1224. (214) von Ossowski, I.; Eaton, J. T.; Czjzek, M.; Perkins, S. J.; Frandsen, T. P.; Schulein, M.; Panine, P.; Henrissat, B.; Receveur-Brechot, V. Protein disorder: conformational distribution of the flexible linker in a chimeric double cellulase. Biophys. J. 2005, 88, 28232832. (215) Bordelon, T.; Montegudo, S. K.; Pakhomova, S.; Oldham, M. L.; Newcomer, M. E. A disorder to order transition accompanies catalysis in retinaldehyde dehydrogenase type II. J. Biol. Chem. 2004, 279, 43085-43091. (216) Iakoucheva, L. M.; Radivojac, P.; Brown, C. J.; O’Connor, T. R.; Sikes, J. G.; Obradovic, Z.; Dunker, A. K. The importance of intrinsic disorder for protein phosphorylation. Nucleic Acids Res. 2004, 32, 1037-1049.

PR060392U

Functional Anthology of Intrinsic Disorder. 1. Biological Processes and

Recommend Documents